Semantic based Personalized Framework for Information Retrieval

Traditional information retrieval systems are mostly keywordbased and retrieve documents or information by matching keywords. These systems lack a meaningful description for information, so it is difficult for users to find more relevant information. To provide what a user really needs, a framework of information retrieval based on semantics has been proposed in this paper. In this framework the semantics in the user query are identified and these are summarized according to the context. Then the results are classified into possible domains or groups and displayed to user according to his choice from domain the results are re-ranked. By this framework we provide users a convenient and more precise search service with personalization. General Terms Information Retrieval, Clustering, Ranking