User Centric Web Search Using Ontology

Semantic information retrieval systems query the World Wide Web based on context information, and are intended to provide more pertinent search results. However, most of the existing systems overlook one important aspect 'the user'. They are more focused on eliminating the obscure results that a conventional or non-semantic search engine would throw up and hence, they are pretty much static. On the other hand, our effort would channel its focus more towards providing a more user-centric service using ontology and involving learning and prediction. By studying the usage statistics of the user, context information can be built and used effectively to produce better search results. Such an approach also entails that the knowledge that is accrued, be organized such that the relationships between the data elements can be elicited easily and unambiguously. The ontology would be described using OWL. Latent semantic indexing algorithm is used for context analysis and retrieval.