Increasing amount of information on the web has motivated the creation of the recommendationsystems. These systems support users that are associated with large information spaces and lead themto the information they exactly need.Recommendation systems are still limited by several problems, such as sparsity, or the new userproblem. They also fail to make full usage and harness the power of domain knowledge and semanticweb ontologies. In this paper a recommendation system, called OntoTimes, is proposed that comprisessemantic relation of users associated tags with site’s ontology, further, it involves users time interestsin recommendations. Evaluations demonstrate that OntoTimes recommendation system, since userstime interests are involved, achieves better accuracy (47%) against the recommendation system thatare only based on ontology (37%).
Manuscript profile
