Personalized web search ranking based on different information resources

Abstract

The goal of personalized search is to provide user with results that accurately satisfy their specific goal of the search. In this paper, a hybrid personalized search re-ranking approach is proposed to provide users with results reordered according to their interests. User preferences are automatically learned into a concept-based user profile. This profile is then employed in the re-ranking process with other information resources to personalize results. Our experiments have shown interesting results in enhancing the quality of web search.