Ontology Based Navigation Pattern Mining For Efficient Web Usage

Users of the web have their own areas of interest. Given the tremendous growth of the web, it is very difficult to redirect the users to their page of interest. Web usage mining techniques can be applied to study the users navigational behaviours based on their previous visit data. These user navigational patterns can be extracted and used for web personalization or web site reorganization recommendations. Web usage mining techniques do not use the semantic knowledge of the web site for such navigational pattern discovery. But, if ontology is applied along with web usage techniques, it can improve the quality of the detected patterns. This research work aims to design a framework that integrates semantic knowledge with web usage mining process that generates the refined website ontology that recommends personalization of web. As the web pages are seen as ontology individuals, the user navigational behaviours over a certain period are considered as the user expected ontology refinement. The user profiles and the web site ontology are compared and the variation between the two is proposed as the new refined web site ontology. The web site ontology has been semi-automatically built and evolves through the adaptation procedure. The result of implementation of this recommendation system indicates that integrating semantic information and page access patterns yield more accurate recommendations. Keyword-Web Usage Mining, Semantic Web, Ontology, Web Page Recommendation