Reconnaissance d'entités nommées : enrichissement d'un système à base de connaissances à partir de techniques de fouille de textes
Résumé
In this paper, we present and analyze the results obtained by our named entity recognition system, CasEN, during the Ester2 evaluation campaign. We identify on what difficulties our system was the most challenged, which mainly are: out-of-vocabulary words, metonymy and detection of the boundaries of named entities. Next, we propose a direction which may help us for improving performances of our system, by using exhaustive hierarchical and sequential data mining algorithms. This approach aims at extracting patterns corresponding to useful linguistic constructs for recognizing named entities. Finaly, we describe our experiments, give the results we currently obtain and analyze those results.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...