An Efficient Framework for Document Retrieval in Relationship Strengthened Personalized Ontologies

Abstract

Personalization is crucial in the prevailing Internet scenario. An efficient information retrieval not only relies on personalization but also upon security. There is need of a semantic approach of document retrieval to boost the search quality. The incorporation of knowledge extracted from multiple ontologies is the significant research direction to get the most appropriate documents from the knowledge base. In this paper, local and global ontologies are used for mining the semantic associations. Multiple ontologies can share a variety of concepts, and automatic merging of them is feasible. Any input query is often matched with the semantic association for finding the relevant words. This is achieved by analyzing the intensified concept relationships in the local repository or local ontology and finding out the possibilities of automation by similarity calculation. A user profiling model is constructed which is trained for personalization. A novel framework has been proposed which provides document retrieval. The results are analyzed using benchmark SWETO ontology. Due to the query property enhancements, the error rate is reduced and the local ontology gets optimized for synchronization with the global Web ontology, thereby enhancing user personalization.