Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with IGNGF neural clustering : Advances in Self-Organizing Maps
Identifieur interne : 000786 ( Main/Exploration ); précédent : 000785; suivant : 000787Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with IGNGF neural clustering : Advances in Self-Organizing Maps
Auteurs : Jean-Charles Lamirel [France] ; Ingrid Falk [France] ; Claire Gardent [France]Source :
- Neurocomputing : (Amsterdam) [ 0925-2312 ] ; 2015.
Descripteurs français
- Pascal (Inist)
- Etiquetage, Classification non supervisée, Analyse syntaxique, Linguistique, Traitement langage, Langage naturel, Grille, Analyse donnée, Intelligence artificielle, Critère sélection, Verbe, Sémantique, Relation syntaxique, Analyse quantitative, Amas, Méthode différentielle, Réseau neuronal, Analyse amas, Analyse qualitative, Algorithme apprentissage, Algorithme en ligne, ., Capabilité objet.
- Wicri :
English descriptors
- KwdEn :
- Artificial intelligence, Cluster, Cluster analysis, Data analysis, Differential method, Grid, Labelling, Language processing, Learning algorithm, Linguistics, Natural language, Neural network, Object-capabilities, Online algorithm, Qualitative analysis, Quantitative analysis, Selection criterion, Semantics, Syntactic analysis, Syntactic relation, Unsupervised classification, Verb.
- mix :
Abstract
Classifications which group together verbs and a set of shared syntactic and semantic properties have proven to be useful in both linguistics and Natural Language Processing tasks. However, most existing approaches for automatically acquiring verb classes fail to associate the verb classes produced with an explicit characterisation of the syntactic and semantic properties shared by the class elements. We propose a novel approach to verb clustering which addresses this shortcoming and permits building verb classifications whose classes group together verbs, subcategorisation frames and thematic grids. Our approach involves the use of a recent neural clustering method called IGNGF (Incremental Growing Neural Gas with Feature maximization). The use of a standard distance measure for determining a winner is replaced in IGNGF by feature maximisation measure relying on the features of the data that are associated with clusters during learning. A main advantage of the method is that maximised features used by IGNGF during learning can also be exploited in a final step for accurately labelling the resulting clusters. In this paper, we exploit IGNGF for the unsupervised classification of French verbs and evaluate the obtained clusters (i.e., verb classes) in two different ways. The first way is a quantitative analysis of the clustering process relying on a usual gold standard and on complementary unbiased clustering quality indexes. The second way is a qualitative analysis of the cluster labelling process. Relying on an adapted gold standard, we evaluate the capacity of the IGNGF clusters labels (i.e., subcategorisation frames and thematic grids) to be exploited for bootstraping a VerbNet-like classification for French. Both analyses clearly highlight the advantages of the approach.
Url:
Affiliations:
- France
- Grand Est, Lorraine (région)
- Nancy
- Centre national de la recherche scientifique, Laboratoire lorrain de recherche en informatique et ses applications, Synalp (Loria), Université de Lorraine
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: 000005
- to stream PascalFrancis, to step Curation: 000998
- to stream PascalFrancis, to step Checkpoint: 000000
- to stream Main, to step Merge: 000775
- to stream Hal, to step Corpus: 002276
- to stream Hal, to step Curation: 002276
- to stream Hal, to step Checkpoint: 000643
- to stream Main, to step Merge: 000661
- to stream Main, to step Curation: 000786
Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with IGNGF neural clustering : Advances in Self-Organizing Maps</title><author><name sortKey="Lamirel, Jean Charles" sort="Lamirel, Jean Charles" uniqKey="Lamirel J" first="Jean-Charles" last="Lamirel">Jean-Charles Lamirel</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Synalp-LORIA</s1><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>Synalp-LORIA</wicri:noRegion><wicri:noRegion>Synalp-LORIA</wicri:noRegion><placeName><settlement type="city">Nancy</settlement><region type="region" nuts="2">Grand Est</region><region type="region" nuts="2">Lorraine (région)</region></placeName><orgName type="team" n="7">Synalp (Loria)</orgName><orgName type="lab">Laboratoire lorrain de recherche en informatique et ses applications</orgName><orgName type="university">Université de Lorraine</orgName><orgName type="EPST">Centre national de la recherche scientifique</orgName></affiliation></author><author><name sortKey="Falk, Ingrid" sort="Falk, Ingrid" uniqKey="Falk I" first="Ingrid" last="Falk">Ingrid Falk</name><affiliation wicri:level="1"><inist:fA14 i1="02"><s1>University of Strasbourg</s1><s3>FRA</s3><sZ>2 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>University of Strasbourg</wicri:noRegion><wicri:noRegion>University of Strasbourg</wicri:noRegion></affiliation></author><author><name sortKey="Gardent, Claire" sort="Gardent, Claire" uniqKey="Gardent C" first="Claire" last="Gardent">Claire Gardent</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Synalp-LORIA</s1><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>Synalp-LORIA</wicri:noRegion><wicri:noRegion>Synalp-LORIA</wicri:noRegion></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">15-0006073</idno><date when="2015">2015</date><idno type="stanalyst">PASCAL 15-0006073 INIST</idno><idno type="RBID">Pascal:15-0006073</idno><idno type="wicri:Area/PascalFrancis/Corpus">000005</idno><idno type="wicri:Area/PascalFrancis/Curation">000998</idno><idno type="wicri:Area/PascalFrancis/Checkpoint">000000</idno><idno type="wicri:explorRef" wicri:stream="PascalFrancis" wicri:step="Checkpoint">000000</idno><idno type="wicri:doubleKey">0925-2312:2015:Lamirel J:federating:clustering:and</idno><idno type="wicri:Area/Main/Merge">000775</idno><idno type="wicri:source">HAL</idno><idno type="RBID">Hal:hal-01074277</idno><idno type="url">https://hal.inria.fr/hal-01074277</idno><idno type="wicri:Area/Hal/Corpus">002276</idno><idno type="wicri:Area/Hal/Curation">002276</idno><idno type="wicri:Area/Hal/Checkpoint">000643</idno><idno type="wicri:explorRef" wicri:stream="Hal" wicri:step="Checkpoint">000643</idno><idno type="wicri:doubleKey">0925-2312:2015:Lamirel J:federating:clustering:and</idno><idno type="wicri:Area/Main/Merge">000661</idno><idno type="wicri:Area/Main/Curation">000786</idno><idno type="wicri:Area/Main/Exploration">000786</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with IGNGF neural clustering : Advances in Self-Organizing Maps</title><author><name sortKey="Lamirel, Jean Charles" sort="Lamirel, Jean Charles" uniqKey="Lamirel J" first="Jean-Charles" last="Lamirel">Jean-Charles Lamirel</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Synalp-LORIA</s1><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>Synalp-LORIA</wicri:noRegion><wicri:noRegion>Synalp-LORIA</wicri:noRegion><placeName><settlement type="city">Nancy</settlement><region type="region" nuts="2">Grand Est</region><region type="region" nuts="2">Lorraine (région)</region></placeName><orgName type="team" n="7">Synalp (Loria)</orgName><orgName type="lab">Laboratoire lorrain de recherche en informatique et ses applications</orgName><orgName type="university">Université de Lorraine</orgName><orgName type="EPST">Centre national de la recherche scientifique</orgName></affiliation></author><author><name sortKey="Falk, Ingrid" sort="Falk, Ingrid" uniqKey="Falk I" first="Ingrid" last="Falk">Ingrid Falk</name><affiliation wicri:level="1"><inist:fA14 i1="02"><s1>University of Strasbourg</s1><s3>FRA</s3><sZ>2 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>University of Strasbourg</wicri:noRegion><wicri:noRegion>University of Strasbourg</wicri:noRegion></affiliation></author><author><name sortKey="Gardent, Claire" sort="Gardent, Claire" uniqKey="Gardent C" first="Claire" last="Gardent">Claire Gardent</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Synalp-LORIA</s1><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><wicri:noRegion>Synalp-LORIA</wicri:noRegion><wicri:noRegion>Synalp-LORIA</wicri:noRegion></affiliation></author></analytic><series><title level="j" type="main">Neurocomputing : (Amsterdam)</title><title level="j" type="abbreviated">Neurocomputing : (Amst.)</title><idno type="ISSN">0925-2312</idno><imprint><date when="2015">2015</date></imprint></series></biblStruct></sourceDesc><seriesStmt><title level="j" type="main">Neurocomputing : (Amsterdam)</title><title level="j" type="abbreviated">Neurocomputing : (Amst.)</title><idno type="ISSN">0925-2312</idno></seriesStmt></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Artificial intelligence</term><term>Cluster</term><term>Cluster analysis</term><term>Data analysis</term><term>Differential method</term><term>Grid</term><term>Labelling</term><term>Language processing</term><term>Learning algorithm</term><term>Linguistics</term><term>Natural language</term><term>Neural network</term><term>Object-capabilities</term><term>Online algorithm</term><term>Qualitative analysis</term><term>Quantitative analysis</term><term>Selection criterion</term><term>Semantics</term><term>Syntactic analysis</term><term>Syntactic relation</term><term>Unsupervised classification</term><term>Verb</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>Etiquetage</term><term>Classification non supervisée</term><term>Analyse syntaxique</term><term>Linguistique</term><term>Traitement langage</term><term>Langage naturel</term><term>Grille</term><term>Analyse donnée</term><term>Intelligence artificielle</term><term>Critère sélection</term><term>Verbe</term><term>Sémantique</term><term>Relation syntaxique</term><term>Analyse quantitative</term><term>Amas</term><term>Méthode différentielle</term><term>Réseau neuronal</term><term>Analyse amas</term><term>Analyse qualitative</term><term>Algorithme apprentissage</term><term>Algorithme en ligne</term><term>.</term><term>Capabilité objet</term></keywords><keywords scheme="Wicri" type="topic" xml:lang="fr"><term>Linguistique</term><term>Intelligence artificielle</term><term>Analyse quantitative</term><term>Analyse qualitative</term></keywords><keywords scheme="mix" xml:lang="en"><term>Cluster labelling</term><term>Clustering</term><term>Incremental learning</term><term>NLP</term><term>Neural networks</term><term>Verb classification</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Classifications which group together verbs and a set of shared syntactic and semantic properties have proven to be useful in both linguistics and Natural Language Processing tasks. However, most existing approaches for automatically acquiring verb classes fail to associate the verb classes produced with an explicit characterisation of the syntactic and semantic properties shared by the class elements. We propose a novel approach to verb clustering which addresses this shortcoming and permits building verb classifications whose classes group together verbs, subcategorisation frames and thematic grids. Our approach involves the use of a recent neural clustering method called IGNGF (Incremental Growing Neural Gas with Feature maximization). The use of a standard distance measure for determining a winner is replaced in IGNGF by feature maximisation measure relying on the features of the data that are associated with clusters during learning. A main advantage of the method is that maximised features used by IGNGF during learning can also be exploited in a final step for accurately labelling the resulting clusters. In this paper, we exploit IGNGF for the unsupervised classification of French verbs and evaluate the obtained clusters (i.e., verb classes) in two different ways. The first way is a quantitative analysis of the clustering process relying on a usual gold standard and on complementary unbiased clustering quality indexes. The second way is a qualitative analysis of the cluster labelling process. Relying on an adapted gold standard, we evaluate the capacity of the IGNGF clusters labels (i.e., subcategorisation frames and thematic grids) to be exploited for bootstraping a VerbNet-like classification for French. Both analyses clearly highlight the advantages of the approach.</div></front></TEI><affiliations><list><country><li>France</li></country><region><li>Grand Est</li><li>Lorraine (région)</li></region><settlement><li>Nancy</li></settlement><orgName><li>Centre national de la recherche scientifique</li><li>Laboratoire lorrain de recherche en informatique et ses applications</li><li>Synalp (Loria)</li><li>Université de Lorraine</li></orgName></list><tree><country name="France"><region name="Grand Est"><name sortKey="Lamirel, Jean Charles" sort="Lamirel, Jean Charles" uniqKey="Lamirel J" first="Jean-Charles" last="Lamirel">Jean-Charles Lamirel</name></region><name sortKey="Falk, Ingrid" sort="Falk, Ingrid" uniqKey="Falk I" first="Ingrid" last="Falk">Ingrid Falk</name><name sortKey="Gardent, Claire" sort="Gardent, Claire" uniqKey="Gardent C" first="Claire" last="Gardent">Claire Gardent</name></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000786 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000786 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Lorraine
|area= InforLorV4
|flux= Main
|étape= Exploration
|type= RBID
|clé= Pascal:15-0006073
|texte= Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with IGNGF neural clustering : Advances in Self-Organizing Maps
}}
| This area was generated with Dilib version V0.6.33. |  |