A session based personalized search using an ontological user profile

摘要: Within the information overload on web and diversity of user interests, it is increasingly difficult for search engines to satisfy needs. Personalized tackles this problem by considering profile during search. This paper describes a personalized approach involving semantic graph-based issued from ontology. User refers interest in specific session defined as sequence related queries. It built using score propagation that activates set semantically concepts maintained same merging scheme. We also define boundary recognition mechanism based tracking changes dominant held relatively new submitted query Kendall rank correlation measure. Then, personalization achieved re-ranking results queries profile. Our experimental evaluation carried out HARD 2003 TREC collection shows our effective.