अमूर्त

Semi-Automatic Enrichment Approach of ?Domain Ontology? by using TALN Tools

Khalid TATANE, Brahim ER-RAHA, Sanaa MOUHIM, Chihab eddine CHERKAOUI

In most of researches, the semi-automatic construction of ontology using texts is generally based on textual medium, and in which the studied domain describes text content. In order to conceive semantically the richest and updated ontology, we suggest developing the classical methods of ontology construction (1), by taking into account the text “content” to construct first kernel of ontology, and (2) by enriching obtained ontology using external resources (public texts, controlled vocabulary of the same domain, specialized web pages, text documents, etc.). This paper describes how these different resources are analyzed and exploited. We have experimented this approach on texts dealing with subjects related to tourism in morocco, and we have assessed the advantage drawn from one richest ontology of the domain (in comparison to a first realized ontology of the domain in the framework of the project GECO-WES1), and we have noted that the results have seen significant improvement.