Approche par réutilisation d'annotations sémantiques pour la recherche d'information sur le web

In this paper, we present a new approach for information retrieval assistance on the web. Its objective is to present to the current user relevant retrieved documents for his query and adapted to his profile. Its consists on using the mechanism of the Case Based Reasoning (CBR) in order to memorize research sessions carried out by users (user profile, query, semantic annotations of relevant documents, date), then to reuse them when a similar research session arises. The reuse annotations during a current session, can also automatically reformulate the current query for improving the quality of responses to the current session. Our proposals have been tested and validated by the development of the system SYRANNOT implemented in Java, using the infrastructure JENA (hp) and using Google. The first experimental evaluations show a significant improvement of results offered by our system compared to google