An ontology-based hybrid recommender system for internet protocol television

Abstract

Internet Protocol Television (IPTV) has gained popularity in providing TV channels and program choices to broad range of user. The service providers are attempting ways to attract more users’ subscription and as from user point of view, they would like to have channel or program recommendations based on their preferences as well as public suggestions. This motivates us to propose an ontology-based hybrid recommender system. This system applies content-based and collaborative filtering in IPTV domain to increase users’ satisfaction. The preliminary experimental results show that our proposed system works more effectively by eliminating the cold-start problem, over specialization, data sparsity and new item problems and efficiently by using the ontological user profile for computation of recommendations.