CIDE (2002) Fellah

De CIDE

ArabTDM: Automatic recognition of arabic contents of scientific journals


 
 

 
titre
ArabTDM: Automatic recognition of arabic contents of scientific journals
Auteurs
FELLAH Meriem (1) ; ANJA HABACHA CHAIBI (1) ; MOHAMED BEN AHMED (1) ;
Affiliations
  • RIADI/ENSI, University of La Manouba, TUNISIE
In
Actes du colloque CIDE.05 (Hammamet 2002)
Source
INIST
Abstract
The work presented in this paper tackling the document analysis and recognition problem, in line with a virtual library project. It presents the design and the implementation of an automatic recognition system of the contents of scientific journals written in Arabic. The system deals with homogeneous and intermediary contents. Its based on an hybrid approach which integrates the top-down approach for the segmentation of the digitised contents into significant blocs and the extraction of the main bloc, and the bottom-up approach which considers the text file representing the body of the contents and uses the tagging technique to extract its different items and to distinguish their relating components: Title, Author(s) and Page numbers. The output of the system is a HTML (HyperText Markup Language) file stored into a database with appropriate hyperlinks.


Mots-clés anglais 
HTML language ; Segmentation ; Content analysis ; Arabic ; Language ; Periodical ; Scientific literature ; Optical character recognition ;
Mots-clés français 
Langage HTML ; Segmentation ; Analyse contenu ; Arabe ; Langage ; Périodique ; Littérature scientifique ; Reconnaissance optique caractère ;
Mots-clés espagnols 
Lenguaje HTML ; Segmentación ; Análisis contenido ; Árabe ; Lenguaje ; Periódico ; Literatura científica ; Reconocimento óptico de caracteres ;
logo travaux Référence à améliorer :
sémantisation des auteurs et affiliations
… davantage au sujet de « CIDE (2002) Fellah »
ArabTDM: Automatic recognition of arabic contents of scientific journals +