Wicri ICS - User contributions [en]

Template:Small

2024-03-31T07:49:36Z

Jacques Ducloy: Created page with "{{{1}}}"

{{{1}}}

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-31T07:49:21Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[{{Article body/Link bibl|2}}].;
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Begin}}
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
 < {{Small|http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html}} >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005 < doi:10.1045/july2005-lynch >
}}

{{Article body/Bibl
|id=11.
|author=M. Patel,
|text=Fourth Open Archives Forum Workshop In Practice, Good Practice: The Future
of Open Archives, Ariadne Issue 37, Oct 2003, < http://www.ariadne.ac.uk/issue37/oa-forum-ws-rpt/#6>
}}

{{Article body/Bibl
|id=12.
|author=M. Smith, M. Bass, G. McClellan, R. Tansley, M. Barton, M. Branschofsky, D. Stuve,
and J. H. Walker,
|text=DSpace: An Open Source Dynamic Digital Repository, D-Lib Magazine, 9
(1), 2003.
<doi:10.1045/january2003-smith>.
}}

{{Article body/Bibl
|id=13.
|author=H. Suleman, A. Atkins, M. Gonçalves, R. France and E. Fox.
|text=Networked Digital Library
of Theses and Dissertations, Bridging the Gaps for Global Access - Part 1: Mission and
Progress. In D-lib Magazine, September 2001
< doi:10.1045/september2001-suleman-pt1>
}}
{{Article body/End}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-31T07:45:13Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[{{Article body/Link bibl|2}}].;
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Begin}}
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
 < http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005 < doi:10.1045/july2005-lynch >
}}

{{Article body/Bibl
|id=11.
|author=M. Patel,
|text=Fourth Open Archives Forum Workshop In Practice, Good Practice: The Future
of Open Archives, Ariadne Issue 37, Oct 2003, < http://www.ariadne.ac.uk/issue37/oa-forum-ws-rpt/#6>
}}

{{Article body/Bibl
|id=12.
|author=M. Smith, M. Bass, G. McClellan, R. Tansley, M. Barton, M. Branschofsky, D. Stuve,
and J. H. Walker,
|text=DSpace: An Open Source Dynamic Digital Repository, D-Lib Magazine, 9
(1), 2003.
<doi:10.1045/january2003-smith>.
}}

{{Article body/Bibl
|id=13.
|author=H. Suleman, A. Atkins, M. Gonçalves, R. France and E. Fox.
|text=Networked Digital Library
of Theses and Dissertations, Bridging the Gaps for Global Access - Part 1: Mission and
Progress. In D-lib Magazine, September 2001
< doi:10.1045/september2001-suleman-pt1>
}}
{{Article body/End}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

DC 2010 Pittsburgh

2024-03-30T23:01:23Z

Jacques Ducloy: /* Program */

[[Image:DC2010Banner.jpg|center]]
{{Semantic Infobox Event
| Title=DC 2010
| Series=DC Conference
| Start date=Oct 20 2010
| Country=United States
| City=Pittsburgh
| Homepage=www.asis.org/Conferences/DC2010/
| Homepage label=www.asis.org/ ... /DC2010
| Paper deadline=Apr 9 2010
}}
<big><center>'''Making Metadata Work Harder: ''Celebrating 15 Years of Dublin Core '''''</center></big>

==Conference Theme==

DC-2010 marks the 15th anniversary of Dublin Core and the 10th year of the Annual Conference. To celebrate and reflect on the past and future, conference participants will engage in investigations in both research and application entailing advances that make metadata work harder in ways beyond the originally identified need for better resource discovery. The DCMI Abstract Model, the refined concept of Application Profiles and the initial Description Set Profile specification, represent an inflection point in the trajectory of metadata design and deployment in the service of human information needs. But all is not perfect and we have much to learn from sharing experience. DC-2010, will take stock of progress, look to the future and celebrate the broad scope of research and applied work in making metadata work harder.

Beyond the conference theme, papers, reports, and poster submissions are welcome on a wide range of metadata topics, such as:

*Metadata principles, guidelines, and best practices
*Metadata quality, normalization, improvement and mapping
*Conceptual models and frameworks (e.g., RDF, DCAM, OAIS)
*Application profiles
*Metadata interoperability across domains, languages, time, structures, and scales.
*Cross-domain metadata uses (e.g., recordkeeping, preservation, curation, institutional repositories, publishing)
*Domain metadata (e.g., for corporations, cultural memory institutions, education, government, and scientific fields)
*Bibliographic standards (e.g., RDA, FRBR, subject headings) as Semantic Web vocabularies
*Accessibility metadata
*Metadata for scientific data, e-Science and grid applications
*Social tagging and user participation in building metadata
*Knowledge Organization Systems (e.g., ontologies, taxonomies, authority files, folksonomies, and thesauri) and Simple Knowledge Organization Systems (SKOS)
* Ontology design and development
* Integration of metadata and ontologies
* Metadata generation (methods, tools, and practices)
* Search engines and metadata
* Semantic Web metadata and applications
* Vocabulary registries and registry services

==Submissions==

Authors wishing to submit papers, reports, or poster proposals may do so through the DCMI Peer Review System at http://dcpapers.dublincore.org/index.php/pubs/. Author registration and instructions for the submission process appear under the “Information for Authors” link. Author Guidelines for full papers, project reports and posters/demonstrations are available through the DCMI Peer Review System. All submissions to the DC-2010 Conference Proceedings will be peer-reviewed by the International Program Committee. All submissions must be in English. Accepted submissions will be published in the official electronic Conference Proceedings. Unless previously arranged, accepted papers, project reports and posters must be presented in Pittsburgh by at least one of their authors.

All submitting authors must provide basic information regarding current professional positions and affiliations as a condition of acceptance and publication.

===Full papers (8-10 pages)===

Full papers either describe innovative work in detail or provide critical, well-referenced overviews of key developments or good practice in the areas outlined above. Full papers will be assessed using the following criteria:

* Originality of the approach to implementation
* Quality of the contribution to the implementation community
* Significance of the results presented
* Clarity of presentation

===Project reports (4-5 pages)===

Project reports describe a specific model, application, or activity in a concise, prescribed format. Project reports will be assessed using the following criteria:

* Conciseness and completeness of technical description
* Usability of the technical description by other potential implementers
* Clarity of presentation

===Posters & demonstrations (1-2 pages)===

Posters are for the presentation of projects or research under development or late-breaking results. Poster proposals should consist of a one-two page extended abstract. Posters will be assessed using the following criteria:

* Concise statement of research or project goals and milestones
* Significance of the research or project
* Framing of key barriers and future research
* Statement of results and accomplishments
* Clarity of presentation

Accepted posters will be published in the Conference Proceedings and displayed at the conference. Unless otherwise arranged, accepted posters must be presented in Pittsburgh by at least one of their authors. However, with prior arrangement, posters may be included in the proceedings and presented by means of video ranging from 4-10 minutes in length and uploaded to YouTube with the link supplied at the time the poster is submitted. Asynchronous mechanisms for participant/author communications will be provided.

==Committees==

===Organization===

* [[Dublin Core Metadata Initiative]] ;
* with [[American Society for Information Science and Technology]] (ASIS&T).

===Comittees===

; Conference Committee Chair:
:*[[Stuart A. Sutton]], [[A pour affiliation de président de comité de programme::Université de Washington]], [[USA]].

;Conference Committee Co-chairs ,:
:* [[Diane Hillmann]], [[A pour affiliation de président de comité de programme::Information Institute of Syracuse ]], [[USA]] ;
:* [[Mike Lauruhn]], [[A pour affiliation de président de comité de programme::Taxonomy Strategies]], [[USA]]
;Pour les workshops :
:* [[Liddy Nevile]], [[Université de La Trobe|La Trobe University]], [[Australie]]
;Pour les tutoriaux :
:* [[Marcia Lei Zeng]], [[Université d'État de Kent|Kent State University]], [[États-Unis]]

;Program Committee:
*[[Has PC member::Leif Andresen]], [[Danish Agency for Libraries and Media]], [[Denmark]]
*[[Has PC member::Ann Apps]], [[BSc MBCS CITP]], Retired, [[United Kingdom]]
*[[Has PC member::Thomas Baker]], [[DCMI Directorate]], [[USA]]
*[[Has PC member::Ana Alice Baptista]], [[Universidade do Minho]], [[Portugal]]
*[[Has PC member::Michael Robert Bolam]], [[University of Pittsburgh]], [[USA]]
*[[Has PC member::Joseph A. Busch]], [[Project Performance Corporation]], [[USA]]
*[[Has PC member::Douglas Campbell]], [[National Library of New Zealand]], [[New Zealand]]
*[[Has PC member::Chao-chen Chen]], [[National Taiwan Normal University]], [[Taiwan]]
*[[Has PC member::Eric Rogers Childress]], [[OCLC Research]], [[USA]]
*[[Has PC member::Michael D. Crandall]], [[University of Washington]], [[USA]]
*[[Has PC member::Jacques Ducloy]], [[DRRT Lorraine]], [[France]]
*[[Has PC member::Schubert Shou Boon Foo]], [[Nanyang Technological University]], [[Singapore]]
*[[Has PC member::Jane Greenberg]], [[University of North Carolina]], Chapel Hill, [[USA]]
*[[Has PC member::Corey A. Harper]], [[New York University]], [[USA]]
*[[Has PC member::Bernhard Haslhofer]], [[University of Vienna]], Austria
*[[Has PC member::Carol Hert]], [[Schema Logic]], Inc. [[USA]]
*[[Has PC member::Gail M. Hodge]], [[Information International Assoc.]], [[USA]]
*[[Has PC member::Pete Johnston]], [[Eduserv Foundation]], United Kingdom
*[[Has PC member::Masahide Kanzaki]], [[Zenon Ltd. Partners]], Japan
*[[Has PC member::Christopher S.G. Khoo]], [[Nanyang Technological University]], Singapore
*[[Has PC member::Traugott Koch]], [[Max Planck Digital Library]], Germany
*[[Has PC member::John Kunze]], [[California Digital Library]], [[USA]]
*[[Has PC member::Faye R. Leibowitz]], [[University of Pittsburgh]], [[USA]]
*[[Has PC member::Yushiana Mansor]], [[International Islamic University Malaysia]], Malaysia
*[[Has PC member::Filiberto Felipe Martinez-Arellano]], [[Universidad Nacional Autónoma de México]], Mexico
*[[Has PC member::Michael Robert]], [[Middleton Queensland University of Technology]], Australia
*[[Has PC member::Alistair Miles]], [[University of Oxford]], United Kingdom
*[[Has PC member::Akira Miyazawa]], [[National Institute of Informatics]], Japan
*[[Has PC member::William E. Moen]], [[University of North Texas]], [[USA]]
*[[Has PC member::Jin-Cheon Na Nanyang]], [[Technological University]], Singapore
*[[Has PC member::Mitsuharu Nagamori]], [[University of Tsukuba]], Japan
*[[Has PC member::Cristina Pattuelli]], [[Pratt Institute]], [[USA]]
*[[Has PC member::Jon Phipps]], [[JES & Co.]], [[USA]]
*[[Has PC member::Jian Qin]], [[Syracuse University]], [[USA]]
*[[Has PC member::Sandy K. Roe]], [[Cataloging & Classification Quarterly]], [[USA]]
*[[Has PC member::Stefanie Ruehle]], [[[[Göttingen State and University Library|SUB Goettingen]]]], Germany
*[[Has PC member::Bernhard Schandl]], [[University of Vienna]], Austria
*[[Has PC member::Aida Slavic]], [[UDC Consortium]], United Kingdom
*[[Has PC member::Shigeo Sugimoto]], [[University of Tsukuba]], Japan
*[[Has PC member::Stuart A. Sutton]], [[University of Washington]], Retired, [[USA]]
*[[Has PC member::Ahmed Taha]], [[UAE University]], United Arab Emirates
*[[Has PC member::Joseph T. Tennis]], [[University of Washington]], [[USA]]
*[[Has PC member::Seth van Hooland]], [[Université Libre de Bruxelles]], Belgium
*[[Has PC member::Sherry L. Vellucci]], [[Rutgers University]], [[USA]]
*[[Has PC member::Cheryl D. Walters]], [[Utah State University]], [[USA]]
*[[Has PC member::Mei-Ling Wang]], [[National Chengchi University]], Taiwan
*[[Has PC member::Hollie C. White]], [[University of North Carolina at Chapel Hill]], [[USA]]
*[[Has PC member::Andrew C. Wilson]], [[Australian National Data Service]], Australia
*[[Has PC member::Marcia Lei Zeng]], [[Kent State University]], [[USA]]

==Program==

;invited speakers:
* [[Has invited speaker::Stu Weibel]]: A Metadata Trajectory: What 15 years of Dublin Core tells us about our future.
* [[Has invited speaker::Mike Bergman]]: Bridging the Gaps: Adaptive Approaches to Data Interoperability.

;Session Principles and Innovations:
* Emma Tonkin, Andrew Hewson: Building blocks of metadata: What can we learn from Lego™?
* Sherry Koshman: Visualizing Metadata for Environmental Datasets

;Session Libraries and the FRBR Model :
*Maja Zumer, [[Marcia Lei Zeng]], Athena Salaba: FRBR: A Generalized Approach to Dublin Core Application Profiles
*Jenn Riley: Enhancing Interoperability of FRBR-Based Metadata
*Jennifer B. Bowen: Moving Library Metadata Toward Linked Data: Opportunities Provided by the eXtensible Catalog

;Session Communities and Metadata :
* [[Marie-Claude Côté]], [[Margaret Devey]], [[Lynne McAvoy]], [[Leigh Bain]] : Celebrating 10 Years of Government of Canada Metadata Standards
* Paul Asman, San Cannon, Christine Sommo: Extending RSS to Meet Central Bank Needs
* Jian Qin, Miao Chen, Xiaozhong Liu, Andrea Kathleen Wiggins: Linking Entities in Scientific Metadata

;Session DC in Practice:
*Carol Jean Godby: From records to streams: Merging library and publisher metadata
*Steven J. Miller: The One-To-One Principle: Challenges in Current Practice
*Hannah Tarver: Better Guidelines, Better Functionality: How Metadata Supports the Cycle of System Improvement at UNT

;Session (reports) Metadata at Work:
Yunyun Shen, Long Xiao, Ying Feng: [https://doi.org/10.23106/dcmi.952109981 Building Metadata Application Framework for Chinese Digital Library: A Case Study of National Digital Library of China]
*Sarah Buchanan : Use of Community Metadata: Public Policy Research in Policy Archive
* Myung-Ja K. Han, Sheila Bair, Jason Lee: Creating Metadata Best Practices for CONTENTdm Users

;Session (reports) New Directions in Metadata:
*Stephanie Ogeneski Christensen, Douglas Donald Dunlop: The Case for Implementing Core Descriptive Embedded Metadata at the Smithsonian
*James E. Andrews, Denise Shereff, Timothy Patrick, Rachel Richesson: The Question about Questions: Is DC a Good Choice to Address the Challenges of Representation of Clinical Research Questions and Value Sets?
*Jacques Ducloy, Thierry Daunois, Muriel Foulonneau, Alice Hermann, Jean-Charles Lamirel, Stéphane Sire, Jean-Pierre Thomesse, Christine Vanoirbeek: [[Metadata for WICRI, a Network of Semantic Wikis for Communities in Research and Innovation]].

;Tutorials:
* Jane Greenberg, Jon Phipps : Dublin Core: History, Key Concepts, and Evolving Context
* Karen Coyle, Ron Daniel : A SAFARI from the Dublin Core to the Semantic Web

;Working groups:
*Libraries Task group, [http://dublincore.org/librarieswiki/ Wiki], [http://www.jiscmail.ac.uk/lists/DC-LIBRARIES.html Mailing list]
*Preservation Community, [http://dublincore.org/groups/preservation/ Web], [http://www.jiscmail.ac.uk/lists/dc-preservation.html Mailing list]
*Registry Community and Task Group, [http://dublincore.org/groups/registry/ Web], [http://wiki.metadataregistry.org/DCMI_Registry_Community#DCMI_Registry_Community_Projects Wiki], [http://www.jiscmail.ac.uk/lists/DC-REGISTRY.html Mailing list]
*Education Community and Task Group, [http://dublincore.org/groups/education/ Web], [http://dublincore.org/educationwiki/ Wiki], [http://www.jiscmail.ac.uk/lists/DC-EDUCATION.html Mailing list]
*DCMI/NKOS Task Group, [http://dublincore.org/groups/nkos/ Web], [http://www.metadataetc.org/wiki/dcmi-nkos/doku.php Wiki], [http://www.jiscmail.ac.uk/lists/DC-NKOS.html Mailing list]
*User Documentation and Glossary Task Group, [http://sites.google.com/site/dublincoreglossaryupdate/ Wiki], [http://www.jiscmail.ac.uk/lists/DC-GLOSSARY.html Mailing list]
*Libraries Community, [http://dublincore.org/groups/libraries/ Web], [http://www.jiscmail.ac.uk/lists/DC-LIBRARIES.html Mailing list]
*Science and Metadata Community, [http://dublincore.org/groups/sam/ Web], [http://www.jiscmail.ac.uk/lists/DC-SCIENCE.html Mailing list]
*Knowledge Management Community, [http://dublincore.org/groups/km/ Web], [http://www.jiscmail.ac.uk/lists/DC-KM.html Mailing list]
*DCMI/RDA Task Group , [http://dublincore.org/dcmirdataskgroup/ Wiki], [http://www.jiscmail.ac.uk/lists/DC-RDA.html Mailing list]
*Tools Community , [http://dublincore.org/groups/tools/ Web], [http://www.jiscmail.ac.uk/lists/DC-TOOLS.html Mailing list]
*Social Tagging Community, [http://dublincore.org/groups/social-tagging/ Web], [http://dublincore.org/taggingwiki/ Wiki], [http://www.jiscmail.ac.uk/lists/DC-SOCIAL-TAGGING.html Mailing list]
*Localization and Internationalization Community and Translation Task Group, [http://dublincore.org/groups/languages/ Web], [http://eiah.org/dcwiki/index.php/Translation_Task_Force Wiki], [http://www.jiscmail.ac.uk/lists/DC-INTERNATIONAL.html Mailing list]

[[Category:Metadata]]
[[fr:DC 2010 Pittsburgh]]
__SHOWFACTBOX__

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:38:50Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[{{Article body/Link bibl|2}}].;
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Begin}}
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005 < doi:10.1045/july2005-lynch >
}}

{{Article body/Bibl
|id=11.
|author=M. Patel,
|text=Fourth Open Archives Forum Workshop In Practice, Good Practice: The Future
of Open Archives, Ariadne Issue 37, Oct 2003, < http://www.ariadne.ac.uk/issue37/oa-forum-ws-rpt/#6>
}}

{{Article body/Bibl
|id=12.
|author=M. Smith, M. Bass, G. McClellan, R. Tansley, M. Barton, M. Branschofsky, D. Stuve,
and J. H. Walker,
|text=DSpace: An Open Source Dynamic Digital Repository, D-Lib Magazine, 9
(1), 2003.
<doi:10.1045/january2003-smith>.
}}

{{Article body/Bibl
|id=13.
|author=H. Suleman, A. Atkins, M. Gonçalves, R. France and E. Fox.
|text=Networked Digital Library
of Theses and Dissertations, Bridging the Gaps for Global Access - Part 1: Mission and
Progress. In D-lib Magazine, September 2001
< doi:10.1045/september2001-suleman-pt1>
}}
{{Article body/End}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:36:00Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[{{Article body/Link bibl|2}}].;
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005 < doi:10.1045/july2005-lynch >
}}

{{Article body/Bibl
|id=11.
|author=M. Patel,
|text=Fourth Open Archives Forum Workshop In Practice, Good Practice: The Future
of Open Archives, Ariadne Issue 37, Oct 2003, < http://www.ariadne.ac.uk/issue37/oa-forum-ws-rpt/#6>
}}
12. M. Smith, M. Bass, G. McClellan, R. Tansley, M. Barton, M. Branschofsky, D. Stuve,
and J. H. Walker, DSpace: An Open Source Dynamic Digital Repository, D-Lib Magazine, 9
(1), 2003.
<doi:10.1045/january2003-smith>.
13. H. Suleman, A. Atkins, M. Gonçalves, R. France and E. Fox. Networked Digital Library
of Theses and Dissertations, Bridging the Gaps for Global Access - Part 1: Mission and
Progress. In D-lib Magazine, September 2001
< doi:10.1045/september2001-suleman-pt1>

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:30:12Z

Jacques Ducloy: /* Thematic survey about biodiversity */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[{{Article body/Link bibl|2}}].;
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:25:00Z

Jacques Ducloy: /* Thematic survey about biodiversity */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
This last case study deals with a more thematic aspect of a research policy survey. We have
chosen to speak about biodiversity which is becoming a strategic issue. For instance, the
European Commission launched BiodiverSA37 which aims at “setting up efficient trans-
national co-operation in the field of biodiversity research funding”.

In this section we will study a topic which is not really on the agenda of this project but which
is considered as very close to its targets: how are distributed the activities of public research
laboratories with regard to the main axis of biodiverSA members? This information is
supposed to be contained in the publications and more specifically in the PhD theses.

In order to illustrate the complexity of the problem, here are the figures on R&D biodiversity
funding in Europe:
*more than one hundred funding agencies[2];
*several programs by agency, so several hundreds of programs;
*several project by program, so several thousands of projects;
*several results, such as theses, reports or articles by project, so ten thousand publications!

BiodiverSA intends to create an inventory of all existing biodiversity research funding
programs which will be implemented in a “metadatabase” (on a CERIF38 basis). Vocabulary
aspects will play a fundamental role. More specifically classification (or taxonomy) tools
must be used with some computational constraints in order to produce a set of indicators.

The BiodivERsA classification scheme is still being designed and it could be composed of
three parts.

* A general scientific component based on ASRC (Australian Standard Research Classification). ASRC is tightly related to the Organisation for Economic Cooperation and Development (OECD) Proposed Standard Practice for Surveys of Research and Experimental Development.
* A specific component dedicated to biodiversity, built with a combination of several existing classifications;
* A complementary indexation based on keywords extracted from the “CBD Controlled Vocabulary”39 .

In this context, how could we handle the main topic of this study and, for example, how build
an indicator based on PhD theses?
A first problem is to feed a CERIF compliant database which could be used by BiodivERsA
with something close to qualified Dublin Core. But the most important issue is the mapping of
the resources in the classification system. We could imagine that a few research laboratories
will use this classification system in order to be visible by funding agencies. In this case, the
indexer needs also to be cautious with the future computational usage of its elements. (This is
not the same that archiving or making browsing easier).

But a very large amount of theses related to some particular aspect of biodiversity will not use
this schema and several terminological adaptations will be requested. They could be quite
easy if the theses are indexed with a well known vocabulary (MeSH for instance).

In the other cases, a document content linguistic analysis should be done. Once again, several
vocabulary oriented resources are needed, and this last sample would illustrate the need of
complementary terminological repositories.

{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:21:44Z

Jacques Ducloy: /* Institutional surveys */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

INIST experience shows that detecting relationships between research communities appears
to be a key point for research policy [3]. About PhD theses, the computation of affiliations of
jury’s members could set up several kinds of interesting indicators, for instance dealing with
“hidden” research communities.

On a technical point of view, the problem is to extract from metadata several homogeneous
items, dealing with people or affiliation. It could be easy in a standardised world; but in
reality, a given institution could appear in a quite large number of different lexical forms.
In this purpose, authority files and terminological tables play an important part in the
normalisation of the bibliographic data before being handled in the computational process.

In a first step, the authority files can be used to establish the correspondence between well
defined items, for example, the names of countries. The technique generally used to establish
the equivalent terms and normalise the data fields containing data which differ in terms of
typography (upper case or lower case, etc.) or flexion (plural, singular), is to find a
convergence to a simpler form, similar to a key with which the given form is associated.

For the majority of indicators, the analytical unit (the object of the study) is a geographical or
institutional entity. Publications are assigned to these entities on the basis of an analysis of the
addresses of the authors. Variations in the way the names of countries are written are
numerically limited. Relating publications to institutions is a much more difficult task which
cannot be achieved by a simple analysis of the addresses of the authors appearing in the
publications. Very often, a wide variety of different lexical forms for a given entry is found.

This presupposes the existence of geographical (postcodes, towns, regions, countries) and
institutional (code for the institution, classification of the organisations by sector, etc.)
authority files.

As far as we look for a merely statistical indicator, with a medium quality, this kind of post
process is sufficient. But if we need precise computations, some very complex situations
could appear.
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]

Figure 4 illustrates a realistic situation in Nancy geographical area, as it was two years ago.
This example would just illustrate the complexity of what we could meet in practice. A single
list of affiliation names is not sufficient and some more sophisticated tools are requested.
At the upper level, we have two government funded research institutions (CNRS and INRIA)
and one university (UHP) 36 .

At the medium level (research unit), Loria is a joint unit from UHP, Inria, CNRS and two
other universities. INRIA Lorraine is the name of the INRIA component in Nancy.
Some years before, CRIN was the acronym for a joint unit between CNRS and Universities
but without INRIA. CRIN does not exist at the present time, but a lot of papers or theses are
indexed by CRIN and must be handled if we need an “historical study”.

At the bottom level (project team), most of the teams, like Cortex, are part of all upper
organisations. But things could be more complicated! For instance, YT (for Young Team) is
only recognised by universities and not by INRIA (and thus by Inria Lorraine). Orpailleur is
getting recognised by INRIA. Oméga is a joint team between Inria Loraine and INRIA Sophia
but not with Loria.

A consistent metadata schema, such as LEAF[7] one, could offer a solid base which must be
completed by a strong study of affiliation links. An ARTIST working group intends to work
on this kind of relationship, by using for instance several links which could be issued from
something like a “taxonomy of affiliations”.

====Thematic survey about biodiversity====
{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:19:19Z

Jacques Ducloy: /* Creating metadata: thinking about reusability */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
As mentioned before, French research institutions encourage researchers to deposit their work
in an open-access repository for greater visibility and added value of their scientific
production. These repositories include various types of documents or data: some published
articles but also expertises, reports, courses, lecture notes, conference papers, thesis, software
documentation or primary data (demographical for instance).

This repository is a significant and valuable source of information for the evaluation of
researchers and their research units based on their scientific production. For instance,
researchers have to provide information about their scientific production to the evaluation
committee every four years. Why do the researchers or their units have to provide this kind of
information already available in the repository?

INRA, like any other public research institution, could be willing to transfer data from
document repository to the application managing the evaluation files of the researchers and
research units. But, even if the metadata provided from the repository is useful, the evaluation
process requires new and more complex metadata.

To simplify the study, let’s look at the most fundamental exportable data which are required
for the “evaluation” application.

;People: the researcher who has completed his/her PhD thesis in a research institution has to be identified by the evaluation committee as a researcher, a former PhD student and author of a thesis available in the repository. The researcher may have changed his/her name. Identifying the various statuses or names of a person in order to establish correspondences requires enriching the metadata related to persons.
;Structures: the research unit where the PhD student worked may be different from the unit he is working for later on as a researcher. Both structures are entitled to claim the search results presented in the PhD thesis, the first one as research work financial support and the second one as researcher’s affiliation. The “evaluation” application needs to identify the structure the way it was mentioned in the PhD theses with its equivalent in the institution structural network
;Partners: the variety of research institutions in France urges to set up a list of scientific partners and to describe the various collaborations. In this way, the PhD student’s enrolment university is mentioned in the thesis. The list of partners and/or the collaboration type will have to be completed.

====Institutional surveys====

...
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]
....
====Thematic survey about biodiversity====
{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:16:22Z

Jacques Ducloy: /* Three case-studies */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), [[Has author::Jean-Paul Ducasse]], [[Has author::Muriel Foulonneau]], [[Has author::Luc Grivel]], [[Has author::Diane Le Hénaff]], [[Has author::Yann Nicolas]]
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)<ref><nowiki><</nowiki> http://www.loc.gov/standards/mods/ ></ref> are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS<ref><nowiki>< </nowiki>http://www.tdl.org/projects/metadata/tdlappprofile.pdf ></ref>. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation <ref><nowiki><</nowiki> http://www.fcla.edu/dlini/etd.html > </ref> , or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum<ref> < http://www.oaforum.org/ > </ref> [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.

The standardisation of the European Research Systems is also supported by the Commission.
For instance, EuroCRIS <ref> < http://www.eurocris.org/en/ > </ref> aims at “transforming research information into knowledge” while
maintaining and publishing the CERIF <ref> < http://www.cordis.lu/cerif/home.html > </ref> (Common European Research Information Format)
recommendation

Nevertheless, the major initiatives to concretely build a framework for scholarly
communication were launched at national level. The JISC (Joint Information Systems
Committee) has funded projects such as Thesis Alive!16 and Daedalus 17 to promote the
electronic publishing of theses and dissertations in the UK and the integration of UK
institutions in the NDLTD network. SURF (higher education and research partnership
organisation for network services and information and communications technology) has
supported DARE (Digital Academic Repositories)18 project to modify the infrastructure of
provision of academic information in the Netherlands. However, similar initiatives to create
comprehensive frameworks for publishing scholarly material at national level do not exist in
all European countries.

The European IST priority on Research Networking (IST 2.5.6) will face the challenge of
building a framework for publishing scholarly material, at European level. The DRIVER
project (2006-2008) coordinated by the University of Athens will help provide this necessary
infrastructure for European research. It will be based on the open infrastructure proposed in
the scope of the DELOS network of Excellence for digital libraries19 .

In practice, European actors have extremely diverse administrative organisations, inherited
from the past. Interoperability between national systems will have to deal with the
heterogeneity of the structures of academic and research entities, their dependencies and
relations (as detailed below in section 4.2). Additionally, the implementation of a European
framework for e-research will have to face the challenge of multilingualism, with particular
impacts on metadata creation and the management of terminologies.

====The francophone context====
Francophone e-research networks also face both organisational and linguistic challenges. For
the most part, francophone countries (more than 50 countries over 5 continents) are outside
Europe. They have extremely different research infrastructures. Several institutions contribute
in structuring this community. For instance, directly related to theses and dissertations, the
“Organisation Internationale de la Francophonie” (OIF)20 has funded Cyberthèses (see section
3.1). Several institutions such as the “Agence universitaire de la Francophonie” (AUF)21 and
programs related to research infrastructures such as “Système d’Information Scientifique et
Technique” (SIST)22 are also helping standardising scholarly publishing in the francophone
area.

Many francophone countries actually use multiple languages. They need to implement
multilingual systems, with classic constraints in the case of Latin languages and more
complex ones in the case of the Arabic language for example. The IMIST (Moroccan Institute
for Scientific and Technical information)23 in Morocco will implement a bilingual union
catalogue for theses and dissertations24 .

====The French context====
The French administrative organisation is particularly complex because of the multiplicity of
complementary administrative frameworks (an example will be given further in section 4.2).
In the last 10 years, no ambitious program has been launched in France to structure scholarly
publishing at national level. Public institutions in charge of libraries and scholarly
communication such as ABES (Association for Libraries in Higher Education)25 and INIST
(Institute for Scientific and Technical Information)26 have essentially initiated operational
projects such as an integrated publishing chain from articles deposit to the extraction of key
indicators for research. Local initiatives are often disconnected from those operations
launched at national level.

As a result, the focus of operations launched by French actors tends to be too narrow to enable
the implementation of a digital library for e-research, which would federate scholarly
communication at national level.

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

Cyberthèses was born within a francophone program which was also extended to South
America. Cyberdocs, its related platform, is an open source software which supports an
assembly line starting from document writing to dissemination and archiving.

The main members of Cyberthèses network in the Francophony are the following:
“Universidad de Chile” in Santiago27 , “Université de Dakar” (Senegal), “Université
d’Antananarivo” (Madagascar) and the National Institute of Agronomy of Algiers [1].

In the Cyberthèses project each university is in charge of the conversion of its theses and
dissertations into an archiving format (e.g. TEI-lite in XML). At “Université de Lyon 2”, the
electronic registration and deposit are now included in the "charte des thèses" which defines
the relationship between the student and the institution. The deposit of a complete electronic
version of the dissertation is compulsory. The registration is still done by administration, but a
workflow software tool was developed which handles the actual deposit and the electronic
management of the document and its metadata (DC28 , ETDMS, OAI-PMH).

====CCSD: open archive with institutional views====

CCSD stands for “Centre de la Communication Scientifique Directe” and aims at promoting
direct scientific communication between researchers. Very close to ArXiv’s philosophy,
HAL's <ref> HAL stands for “hyper article en ligne” < http://hal.ccsd.cnrs.fr/?langue=en > </ref> software provides an interface for authors to upload into the CCSD database their
manuscripts of scholarly articles in all fields. Most of the French research organisations have
set up a global agreement for a common cooperation based on HAL which can offer an
institutional view for any participant.

A specific service called TEL (thèses-EN-ligne) is dedicated to facilitating the self archiving
of thesis manuscripts, which are important documents for direct scientific communication
between scientists. TEL can be harvested through the OAI-PHM protocol and two metadata
formats are available: unqualified Dublin Core30 , and a specific CCSD one.
A particular feature of this format deals with formal and precise relationships between authors
and affiliations which are clearly identified in deposit procedure. This facility allows the
institutional views and illustrates the two main goals of CCSD: open archive with a research
management orientation.

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.

In the output process, the digital theses will be delivered to a national digital preservation
system which is handled by CINES (Centre Informatique National de l'Enseignement
Supérieur)31 . In addition, metadata will be converted to UNIMARC in order to be sent to
Sudoc union catalogue which hosts the theses national bibliography.

Several complementary services (figure 1) will be offered to institutions of PhD defence:
* Sending to CCSD/HAL and other bibliographic databases;
* Full text indexing in SUDOC 32 (Système Universitaire de DOCumentation) academic portal;
* Building a permanent identifier (URI) and resolution for guaranteeing access in any location to a valid copy of the thesis.
Thus, through a unique deposit, a local institution will be able to provide long term
preservation and dissemination by many channels, with a high level of traceability in both
scientific and administrative aspects.

STAR does not claim to dispense with specific tools or workflows set up by universities. It is
true that STAR will offer a web interface to those universities that don't possess any local
ETDs management tool. For the others, STAR will ingest locally generated metadata and
document files. These metadata will comply with the French exchange format TEF.

''TEF (Thèses Electroniques Françaises)'' is a recommendation provided by an AFNOR33
working group (AFNOR CG46/CN357/GE5). It aims at offering a coherent and flexible
organisation for rich and normalised theses metadata: bibliographic metadata (DC), rights
metadata (METS Rights), administrative metadata relative to the diploma and preservation
metadata. Within TEF, FRBR 34 (Functional Requirements for Bibliographic Records) model
is used as a conceptual tool to untangle the notion of theses, METS as an XML wrapping to
bind the various metadata modules, Schematron35 as a precise and flexible validation tool to
enforce the business rules that come from the French context.

STAR, as a tool, like TEF, as a data structure, plays as a go-between for the benefit of those
that produce and authenticate the theses and their metadata as well as for those that make use
of them.

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]

====INIST: Metadata homogenisation====
INIST (INstitut de l’Information Scientifique et Technique) is a documentary centre which
produces bibliographic databases (Pascal and Francis).

This activity is in permanent evolution. Until fifteen years ago, bibliographic records were
manually produced in IS0 2709 format. In a first step, an equivalent SGML DTD was used in
order to modernise the production process. Now INIST aims at metadata homogenisation
towards a Dublin Core compliant xml schema (Exodic) with automatic indexing.

[[File:Wicri DC 2006 fig 2.jpg|center|300px|thumb|Figure 2]]

One of INIST’s departments is specialised in building thematic portal or handling statistical
studies dealing with research policies (Figure 2). This entity is more and more implied in
defining institutional indicators, bringing INIST, like CCSD, to improve the quality of
metadata related to relationships between authors and affiliations.
{{End article body}}

===Three case-studies===
{{Article body/Begin}}
We have just introduced a set of operators which seems to offer a complete set of library
oriented services. But, for historical reasons, they had been created quite independently. Thus
the reality could be “less than perfect”. This paper is written by several people, coming from
these organisations, who have realized that interoperability was an important issue, and who
are working on exchange formats, generally based on qualified Dublin Core. Is this approach
sufficient?

We will now present several case-studies in which we go beyond the basic bibliographic
needs (deposit and retrieval) in order to introduce some research policy oriented needs.

In the first case we will analyse the handling of a PhD thesis from the start until its
accessibility via OAI-PMH. The “previous designed” workflow shown in Figure 3 looks
simple: a thesis is managed by Cyberthèses, and then sent to STAR, and at last to CCSD to be
integrated within articles flow. We will consider a situation in which two initial organisations,
university and research institution (EPST, Etablissements Publics d’Enseignement et de
Recherche) are concerned.

[[File:Wicri DC 2006 fig 3.jpg|center|400px|thumb|Figure 3]]

Then we will study two cases of using metadata for doing institutional surveys or creating
thematic portals. We will suppose that metadata could have been indexed or upgraded by a
documentary centre such as INIST

====Creating metadata: thinking about reusability====
...
====Institutional surveys====

...
[[File:Wicri DC 2006 fig 4.jpg|center|400px|thumb|Figure 3]]
....
====Thematic survey about biodiversity====
{{End article body}}

===Conclusion===
{{Article body/Begin}}
For this new experience (the writing of this paper), after the translation of “''What Is a Digital Library anyway, anymore''”, we have chosen to work again from a quite technical point of
view. We have identified a large set of stuff40 , such as theses metadata, affiliation links,
vocabulary items, which could upgrade our services. We have underscored the fundamental
role of a set of repositories of various items and naming conventions which should complete
the classical bibliographic archives…

But “''what do we really want to do anyway, anymore?''”

Our common objective is to go further in the e-research or e-science movement and to
consider scientific and technical information regardless of the global needs of the research
organisations. As we are working in separate institutions which manage different objectives or
priorities, this job was not an easy one. Perhaps our most interesting result concerns the
identification of all compromises that we have to work with:
# Compromise between the national environment of theses and the international network.
# Compromise between the different practices of various actors to ensure reusability of metadata through many applications.
# Compromise between the needs specific to every kind of users: librarians, informetrics engineers, policy actors, social aspects in networked collaborations (with a particular point about evaluation: indeed the thesis status guarantees a validation process which is the last step of semantic web).
# Compromise between a focused look on theses and their integration in a larger environment which goes beyond the basic role of a library, even with a “digital” attribute.

In summary, we would consider the theses as nodes within a constellation containing
“articles, dissertations, affiliations, vocabularies”, but also “patents, projects and numerical
results”; in other words all components of a CRIS (Current Research Information System) [8].
Because of a current lack of French or francophone federative research programme, such as
NSDL, ARTIST is trying to set up a place where field actors could experiment and exchange
information about new practices in producing Scientific or Technical Information.
We would like to consider this paper as a step towards a more regular activity. At present step
ARTIST’s services look like a “collective scientific blog” and now we intend to produce a
francophone electronic journal with peer review mechanism, “electronic style” and
sophisticated standardisation. The French language is not to be considered as a “limitation”
and we think that new concepts must be grown deeper in a native language training area
before international confrontation.

In this context, metadata experimentations give us a natural workshop for collaborative
activities that we intend to carry on in the framework of DCMI
{{End article body}}

===Acknowledgements===
{{Article body/Begin}}
Only the main contributors are listed as authors of this paper. We would like to acknowledge
several other people who have contributed by giving some advice or information (Francis
ANDRE, Catherine MOREL-PAIR, Clotilde ROUSSEL and Pierrette PAILLASSARD from
INIST; Amos DAVID from LORIA, Daniel CHARNAY from CCSD; Ghalia MRAHI from
IMIST; Estelle BALIAN from “BiodivErSA – Belgium Biodiversity Platform”, or helping in
the translation or revising process (Marc RUBIO and Catherine GUNET from INIST).
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:15:08Z

Jacques Ducloy: /* INIST: Metadata homogenisation */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:13:47Z

Jacques Ducloy: /* STAR: logistic intermediary between local actors and wider actors. */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:12:44Z

Jacques Ducloy: /* STAR: logistic intermediary between local actors and wider actors. */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:09:22Z

Jacques Ducloy: /* Three case-studies */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:06:07Z

Jacques Ducloy: /* Three case-studies */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T17:04:40Z

Jacques Ducloy: /* INIST: Metadata homogenisation */

File:Wicri DC 2006 fig 4.jpg

2024-03-30T17:01:55Z

Jacques Ducloy:

File:Wicri DC 2006 fig 3.jpg

2024-03-30T17:01:24Z

Jacques Ducloy:

File:Wicri DC 2006 fig 2.jpg

2024-03-30T17:00:58Z

Jacques Ducloy:

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:54:02Z

Jacques Ducloy: /* CCSD: open archive with institutional views */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:52:51Z

Jacques Ducloy: /* CCSD: open archive with institutional views */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:51:55Z

Jacques Ducloy: /* Cyberthèses */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:50:42Z

Jacques Ducloy: /* The French context */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:49:58Z

Jacques Ducloy: /* The francophone context */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:48:55Z

Jacques Ducloy: /* The European context */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:47:14Z

Jacques Ducloy: /* The European context */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:43:03Z

Jacques Ducloy: /* International context of e-research */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:40:13Z

Jacques Ducloy: /* International context of e-research */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:31:41Z

Jacques Ducloy: /* INIST: Metadata homogenisation */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:28:59Z

Jacques Ducloy: /* Conclusion */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:26:16Z

Jacques Ducloy: /* Conclusion */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:25:14Z

Jacques Ducloy: /* Conclusion */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:23:01Z

Jacques Ducloy: /* Several structuring initiatives */

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T16:17:45Z

Jacques Ducloy:

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T15:12:15Z

Jacques Ducloy:

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), Jean-Paul Ducasse, [[Has author::Muriel Foulonneau]], Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.
...

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T15:11:04Z

Jacques Ducloy:

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i), Jean-Paul Ducasse, [[Has author|Muriel Foulonneau]], Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.
...

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T15:09:45Z

Jacques Ducloy: /* STAR: logistic intermediary between local actors and wider actors. */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
From 2006, the French Ministry of Education, which is responsible for PhD theses
infrastructure, will ask ABES, its bibliographic agency, to set up STAR (Signalement des
Thèses, Archivage et Recherche), a new service which will operate as a clearing house.

In the input process, STAR will get theses and related metadata from the institutions entitled
to guarantee that the given document is true to the original which has been validated by the
jury.
...

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-30T15:09:05Z

Jacques Ducloy: /* STAR: logistic intermediary between local actors and wider actors. */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

[[File:Wicri DC 2006 fig 1.jpg|center|300px|thumb|Figure 1]]
{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

File:Wicri DC 2006 fig 1.jpg

2024-03-30T15:07:24Z

Jacques Ducloy:

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:22:41Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=Clifford Lynch{{!}}C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In ''D-lib Magazine'', July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:20:36Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=10
|author=C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In D-lib Magazine, July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:20:10Z

Jacques Ducloy: /* References */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

{{Article body/Bibl
|id=9
|author=C. Lynch.
|text=Where Do We Go From Here? The Next Decade for Digital Libraries. In D-lib Magazine, July 2005
}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:14:18Z

Jacques Ducloy: /* The European context */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====
The European IST (Information Society Technologies) program, like the DLI programs in the
US, has focused on the research dimension of information technologies to create an open
digital library framework. Several projects, such as the Open Archives Forum13 [11], have
been funded by the European Commission to raise awareness of national players and to
investigate the technology issues related to scholarly communication.
...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:13:27Z

Jacques Ducloy: /* International context of e-research */

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University <ref><nowiki><</nowiki> http://publications.uu.se/theses/index.xsql?lang=en></ref>).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====

...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T23:11:07Z

Jacques Ducloy:

{{Wicri working area|text=page in building step}}
[[File:Dcmi-952108464.pdf|200px|right]]
;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University12).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====

...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

File:Dcmi-952108464.pdf

2024-03-29T23:09:53Z

Jacques Ducloy:

Metadata towards an e-research cyberinfrastructure: The case of French PhD theses

2024-03-29T22:58:54Z

Jacques Ducloy: /* International context of e-research */

{{Wicri working area|text=page in building step}}

;Authors:[[Has first creator::Jacques Ducloy]](i),, Jean-Paul Ducasse, Muriel Foulonneau, Luc Grivel, Diane Le Hénaff, Yann Nicolas
* (i) INIST / CNRS
;Abstract: This paper analyses metadata practices and needs in the French research community. It focuses on PhD theses whose life-cycle is totally controlled by the academic institutions. It uses information treatments dealing with setting up research policy as samples for an eresearch orientation. Several case-studies illustrate the fundamental role of various repositories containing affiliations, authorities or linguistic items. ARTIST, the collective author of this paper, is introduced.
==Archived version==
===Introduction===
{{Begin article body}}
This paper is the result of a collaborative work and was written by a networked team of
people, engineers or librarians, working in different organisations, in the framework of
ARTIST<ref>< http://artist.inist.fr/ ></ref> (Appropriation par la Recherche des Technologies de l’Information Scientifique et
Technique) project.
Our first experience was based on various contributions on a
terminological forum, about a translation<ref>< http://artist.inist.fr/article.php3?id_article=245 ></ref> of “''What Is a Digital Library anyway, anymore''”, a
paper written by Carl Lagoze, and whose subject deals with the deep structure of a Digital
Library[{{Article body/Link bibl|9}}]. This paper is a new cooperative experience which would like to analyse how
metadata could help the French academic community in building a federative Digital Library.

The annual issue of the “Academic Ranking of World Universities” [{{Article body/Link bibl|6}}] is causing discomfort
in those in charge of setting up research policies. Improving the quality of metadata items
such as affiliations is now considered as a key issue for improving the visibility of
universities. The researchers themselves are now permanently looking at impact factor. The
“publish or perish” notion is now used as a strong incentive for author self-archiving in
institutional repositories [{{Article body/Link bibl|4}}].

Academic librarian and research communities begin to feel that metadata are not only useful
for information retrieval but could play a more strategic function. This new way of viewing is
perhaps a first step towards a more global analysis about the role of scholarly publishing in
what is called “cyberinfrastructure for e-science or e-research” [{{Article body/Link bibl|10}}].

In this context, this paper will explore how metadata could be used in some activities dealing
with research policy in a francophone<ref>From the French speaking area</ref> environment. We have chosen to focus on PhD theses
because their life-cycle is fully controlled by academic institutions; but a large part of the
discussion could be applied to all items of scholarly publishing.

We will show that a precise research policy requires sophisticated metadata. In an open
archiving framework, the most popular among technical solutions, such as DSpace [{{Article body/Link bibl|12}}], or
Eprints<ref>< http://www.eprints.org/ ></ref>, do not require a depositor to provide strongly structured metadata. Most
requirements are limited to a basic set of Dublin Core elements in order to be easily harvested.
PhD theses are naturally concerned by this goal of improving visibility [{{Article body/Link bibl|5}}]. We will show that
their initial life-cycle requires that metadata should not be merely descriptive but should
include some management elements. Indeed, most of the time and more specifically in a
French context, several institutions or organisations are concerned and must cooperate.

As for all published items of research, theses metadata must be usable in any portal (national,
international, thematic…) that could increase their visibility. They should also be easily
handled by informetric tools in order to be picked out in a scientific or strategic watch or for
research policy oriented studies. At this level we will show that a key issue is the handling of
vocabularies and affiliations.

In the first part of this paper, we will start by introducing the francophone environment. Then
we will present several structuring initiatives dealing with PhD thesis production, union
catalogues and institutional archives. Finally, we will discuss three case-studies showing
various aspects of metadata and vocabularies.
{{Article body/End}}

===Digital libraries for e-research: an overview of European, francophone and French contexts===
{{Article body/Begin}}
Francophone research institutions must position themselves in relation to a variety of existing
national and international frameworks.

They take part in international standardisation initiatives. They have to take into account the
evolution of standards and practices in the United States and worldwide. Additionally, they
are part of both linguistic and regional networks. France and Belgium for instance are part of
both Europe and the francophone area (Francophony). Algeria, Morocco and Tunisia are part
of Francophony as well as of the Arabic language community.

As a result, francophone research actors must coordinate with a number of initiatives in
multiple areas of cooperation. The metadata strategies adopted for scholarly publishing must
ensure interoperability of francophone scholarly material in all those networks. They must
reflect very diverse administrative situations in the different countries as well as in the
regional and international network infrastructures.
====International context of e-research====
The open access movement and the Open Archives Initiative have encouraged research
institutions to make available theses and dissertations on the Web. On the technical side, the
Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)<ref>< http://openarchives.org ></ref> makes it possible to
share and exchange metadata about scholarly material. This has allowed the creation of an
open framework for publishing theses and dissertations. They are integrated into open
repositories and shared in larger networks. In France, this led to the creation of the Centre for
Direct Scholarly Communication<ref>< http://www.ccsd.cnrs.fr/accueil.php3 ?lang=en ></ref> (CCSD), a major initiative aiming at reengineering the
processes of scholarly communication, as illustrated below ([[#CCSD: open archive with institutional views|section 3.2]]).

In the United States, efforts to create an open digital library framework in the scope of the
Digital Library Initiative DLI-I and DLI-II funded by the National Science Foundation have
led to such major projects as the National Science Digital Library<ref>< http://www.nsdl.org ></ref>. NSDL has contributed to
the promotion of standards and the development of services based on an open architecture for
digital libraries. The Networked Digital Library of Theses and Dissertations (NDLTD)<ref><http://www.ndltd.org/index.en.html ></ref> [13]
has developed an infrastructure, including processes and workflow for electronic publishing
of theses and dissertations. It has raised IPR issues related to ETD (electronic theses and
dissertations) publishing. It has also improved repositories technical interoperability by
encouraging the use of OAI-PMH and SRU servers. Finally it has improved metadata-related
interoperability by adopting the ETD metadata set (ETDMS) [4] developed as a Dublin Core
application profile. ETDMS is notably used in the Cyberthèses project (francophone portal for
ETD) further described in section 3.1. Alternative metadata formats such as MARC and
MODS (Metadata Object Description Schema maintained by the Library of Congress)9 are
also used. The Metadata Working Group of the Texas Digital Library has developed a
descriptive application profile for electronic theses and dissertations in MODS10. Finally, a
number of libraries embed descriptive metadata in METS wrappers (e.g. The Florida Center
for Library Automation11, or Uppsala University12).

The ARTIST project, collective author of the present article, is notably in charge of tracking information on the multiplicity of existing metadata initiatives and their evolution in order to ensure that French and francophone actors benefit from those initiatives. It aims to better coordinate the standardisation efforts in the different networks.

====The European context====

...

====The francophone context====

...

====The French context====
...

{{End article body}}

===Several structuring initiatives===
{{Article body/Begin}}
====Cyberthèses====

...

====CCSD: open archive with institutional views====

...

====STAR: logistic intermediary between local actors and wider actors.====
...

{{End article body}}

===References===
{{Article body/Bibl
|id= 1
|text=Y. Bakelli and S. Benrahmoun. Long-term preservation of ETDs in Algeria: discussion
through the CERIST Deposit system. In Proceedings of ETD2003. Berlin 2003.
<http://edoc.hu-berlin.de/conferences/etd2003/bakelli-yahia/HTML/bakelli.html >}}
{{Article body/Bibl
|id= 2
|text= BiodivERsA. Compendium of Biodiversity Research Funding Agencies in Europe
<http://www.eurobiodiversa.org/rich_files/attachments/Compendium%201%20Feb%202006r
ev.doc >}}

{{Article body/Bibl
|id= 3
|author=Luc Grivel{{!}}L. Grivel
|text=H. Fagherazzi, P. Fourneret and A. Zerouki. La conception de bases de
données infométriques hybrides : analyse de la pratique de trois observatoires européens. In
Journées SFBA proceedings Ile Rousse 99.
< http://archivesic.ccsd.cnrs.fr/sic_00000464.html>}}

{{Article body/Bibl
|id= 4
|text=S. Harnad. Publish or Perish - Self-Archive to Flourish : The Green Route to Open
Access. In ERCIM News January 2006
<http://www.ercim.org/publication/Ercim_News/enw64/harnad.html>}}

{{Article body/Bibl
|id= 5
|author=Diane Le Hénaff{{!}}D. Le Henaff
|text= and C. Thiolon. Gérer et diffuser des thèses électroniques : un choix
politique pour un enjeu scientifique. In Documentaliste - Sciences de l’information. 42(4-
5):272-280. October 2005.}}

{{Article body/Bibl
|id= 6
|text= Institute of Higher Education . Academic Ranking of World Universities - Shanghai Jiao
Tong University, 2005
< http://ed.sjtu.edu.cn/ranking.htm >}}
{{Article body/Bibl
|id= 7
|text=M. Kaiser. New Ways of Sharing and Using Authority Information. In D-lib Magazine,
September 2001
<http://www.dlib.org/dlib/november03/lieder/11lieder.html> }}

{{Article body/Bibl
|id=8
|text=K. Jeffery. CRIS + open access = the route to research knowledge on the GRID. In 71st IFLA General Conference and Council proceedings, Oslo, Norway, 2005 <http://www.ifla.org/IV/ifla71/papers/007e-Jeffery.pdf>}}

{{Article body/Bibl
|id=9
|author=Carl Lagoze{{!}}C. Lagoze
|author 2= D. Krafft
|author 3 = S. Payette
|author 4 = S. Jesuroga
|text=What Is a Digital Library anyway, anymore? In ''D-lib Magazine''. November 2005. <http://dx.doi.org/10.1045/november2005-lagoze>}}

===Notes===
<references/>

==See also==

: https://dcpapers.dublincore.org/pubs/article/view/846

__SHOWFACTBOX__

[[fr:Métadonnées dans le contexte d’une cyberinfrastructure de la Recherche]]

Main Page

2021-01-05T15:06:47Z

Jacques Ducloy:

{{Wicri box
|text=<center>'''<big>Welcome on Wicri/Information & Communication Sciences</big> '''See also: [[Artist.en:Main Page|Artist]]</center>
}}
{|width=100%
|-
|[[Image:604px-Julia set 10th power.png|left|140px|:fr:Fractale|Fractale de Julia]]
|<big>'''TICRI'''</big>

* <big>'''T'''</big> Technologies for
* <big>'''I'''</big> Information used by
* <big>'''C'''</big> Communities in
* <big>'''R'''</big> Research
* <big>'''I'''</big> and Innovation
|
{|style="background:#ccccff;-moz-border-radius:0.8em; -webkit-border-radius:0.5em;" align=right
|-
|width=200px|{{Wicri global map
|map size=280
|size 1 = 32
|size 2 = 26
|size 3 = 24
|size 4 = 22
|size 5 = 18
}}
|-
|width=200px|{{Wicri network navbox}}
|}
|}
{{clr}}
{{Begin 2 columns}}
{{H2 Wicri|The wiki Wicri/Ticri (en)}}
{{To French box
|text=This wiki is in the first stage of its building process. To enter the [[wicri.fr:Accueil|wiki Ticri in French]].
}}
{{Skip 2 columns}}
{{DCMI Information}}
{{End 2 columns}}

At this time, this wiki is only dedicated to French speaking communities working on this area in order to define an ontology.

==Work in progress==

Implementing semantic model by an adaptation of [[semweb:Template:Event|Semantic Web]].

* [[Research events]]
* [[Scientific journals]]
* [[Digital information actors]]

[[fr:Accueil]]

==Setting up the wiki==
* [[Projects:Setting up the wiki]]

Main Page

2021-01-05T15:06:00Z

Jacques Ducloy:

{{Wicri box
|text=<center>'''<big>Welcome on Wicri/Open Science</big> '''See also: [[Artist.en:Main Page|Artist]]</center>
}}
{|width=100%
|-
|[[Image:604px-Julia set 10th power.png|left|140px|:fr:Fractale|Fractale de Julia]]
|<big>'''TICRI'''</big>

* <big>'''T'''</big> Technologies for
* <big>'''I'''</big> Information used by
* <big>'''C'''</big> Communities in
* <big>'''R'''</big> Research
* <big>'''I'''</big> and Innovation
|
{|style="background:#ccccff;-moz-border-radius:0.8em; -webkit-border-radius:0.5em;" align=right
|-
|width=200px|{{Wicri global map
|map size=280
|size 1 = 32
|size 2 = 26
|size 3 = 24
|size 4 = 22
|size 5 = 18
}}
|-
|width=200px|{{Wicri network navbox}}
|}
|}
{{clr}}
{{Begin 2 columns}}
{{H2 Wicri|The wiki Wicri/Ticri (en)}}
{{To French box
|text=This wiki is in the first stage of its building process. To enter the [[wicri.fr:Accueil|wiki Ticri in French]].
}}
{{Skip 2 columns}}
{{DCMI Information}}
{{End 2 columns}}

At this time, this wiki is only dedicated to French speaking communities working on this area in order to define an ontology.

==Work in progress==

Implementing semantic model by an adaptation of [[semweb:Template:Event|Semantic Web]].

* [[Research events]]
* [[Scientific journals]]
* [[Digital information actors]]

[[fr:Accueil]]

==Setting up the wiki==
* [[Projects:Setting up the wiki]]