Semantic Information Usage Mining for Next Page Prediction Using Markov Model

Patterns generated by conventional Web Usage Mining methods do not provide explicit insight into the user’s underlying interest and preferences. Hence there is a need to incorporate semantic information in web usage model to understand web user’s navigational behavior at conceptual level. This motivated us to propose the semantically enriched web usage model. The proposed work integrates domain knowledge in the form of ontology in all the phases of Web Usage Mining process. The generated patterns are in terms of ontology instances instead of Web page addresses. Such patterns can extract the semantic relatedness between the visited Web pages. The discovered Semantic rich sequential association rules form the core knowledge of the recommendation engine of the proposed model. Compared with the on the world wide web by extracting conventional Web usage based recommendation system, our proposed model shows promising results. The present work can be extended in the following directions, first elaborating the use of semantically enhanced patterns for automatic or semi automatic Web site adaptation and link restructuring.Second incorporating domain knowledge of the web application in the form of ontology in pattern growth sequential pattern mining algorithms. KeywordsWeb Usage Mining, Semantic Web, Domain Ontology, Sequential Pattern Mining, Markov Model ,Association Rule, Recommender Systems.