QuerySem: Deriving Query Semantics Based on Multiple Ontologies

Internet search engines are indispensable tools that assist users to find information on the World Wide Web (WWW). These search engines use different keyword-based indexing techniques to index Web Pages. Although this approach assist users in finding information on the Web, many of the returned results are irrelevant to the user’s information needs. This is because of the “semantic-gap” between the meanings of the keywords that are used to index Web Pages and the meanings of the terms used by the user to formulate his query. In this paper, we introduce an approach to employ knowledge represented by multiple large-scale general-purpose ontologies to derive the semantic aspects of the user’s query. In addition, we utilize statistical-based semantic relatedness measures to compensate for missing background knowledge in the exploited ontologies. Experimental instantiation of the proposed system validates our proposal.