International Journal of Engineering and Technology

Ontological way of knowledge representation is very much useful to the semantic web. In the modernized computer era, there is a need of a special technique for personalization. XML plays an important role in information retrieval systems and XML being a common format for information interpretation, it will be easy to understand as well as easy to construct. In this paper, a framework has been proposed for personalizing the web using XML based ontologies. This framework needs integration between global ontology and locally generated ontology based on user profiles. The relevant concepts between both the ontologies are identified, grouped together and ranked. Finally, the generated ontologies are evaluated using standard datasets, based on their semantic structures. The clustered concepts and query pairs are being analyzed with varying threshold limits. In addition, the performance metrics show that the ontology based techniques show a good precision, recall values for the user given data, when compared to text-based approaches.