| Paper Details: | Downloads: 428 |
| Serial Number: | P1151052919
|
| Title: | A generic model with radial basis for the indexing of the Arab documents based on the semantic graphs
|
| Authors: | T. ZAKI and D. MAMMASS and A. ENNAJI and S. NICOLAS
|
| Abstract: | {In this paper, we proposed a generic system for the extraction of the elements index that appear within the multilingual documents by extending the vectorial model of Salton by a combination of the calculation of the TFIDF with the formula of Okappi in order to identify the
relevant terms which characterize best document. Indeed, we have proposed a new measure Okappi- TFIDF-ABR which takes in consideration the concept of semantic vicinity using a measure of similarity between terms by combining the calculation of TF-IDF-Okapi with a kernel approach (Radial Basis function). Afin d'avoir une bonne indexation, nous nous sommes basés sur les graphes sémantiques pour mettre en évidence les connexions sémantiques entre termes, ainsi nous avons utilisé un dictionnaire auxiliaire pour augmenter la connexité du graphe ainsi construit et par conséquent le poids sémantique des termes de l'index. This indexing therefore provides a contextual and semantic research, and our system showed a better robustness especially for the documents written in Arab language
|
| Keywords: | Document indexation, semantic graph, semantic vicinity, dictionary, kernel function, okapi formula, similarity, TF-IDF, vectorial model
|
| Journal/Conference: | International Journal of Graphics, Vision and Image Processing
|
| Volume: | 10
|
| Issue: | 6
|
| Submission Date: | 9/1/2010 12:00:00 AM
|
| Review Date: | 10/1/2010 12:00:00 AM
|
| Publishing Date: | 12/31/2010 12:00:00 AM
|
| Article Downloads: | 428
|
| Download: |
|