Ontology Discovery from Web Pages : Application to Tourism
Résumé
This paper deals with the automation of ontology building process from HTML pages. Our methodology is based on the complementary use of two approaches. The first approach is based on Aussenac-Gilles methodology and requires feedback from the user to propose and refine concepts and their relationships. The second one exploits Web pages structure and is based on clustering techniques to group similar words into clusters in order to define a concept hierarchy. We apply our methodology to tourism domain.