Evaluation des performances d'un système de recherche d'information utilisant un algorithme de segmentation thématique de pages Web

In this paper, we explore the use of new page segmentation algorithm using both visual and structural mark-up (,) to partition web pages into blocks and investigate how to take advantage of block-level evidence to improve retrieval performance in the web. We propose a new ranking function that combines content and link rank based on propagation of scores over links on block-to-page graph. This function propagates scores from blocks of source pages to destination pages in relation with query terms. Our approach shows good results over TREC collections. MOTS-CLÉS : recherche d’information, systèmes hypertextes, analyse de liens, web, propagation de pertinence, méthodes de segmentation, analyse thématique.