Personalized TV contents recommender system using collaborative context tagging-based user's preference prediction technique

Abstract

Increasing amounts of TV contents are being disseminated through advanced broadcast, satellite and smart TV technologies leading to an information overload. Despite the excessive increase of TV contents, there are few considerations in the development of novel methods for personalized smart TV content recommender services. Most existing personalized TV content services are mainly focus on using individual profiles and require user's immediate participation in rating their experienced contents. In this paper, we propose the context tagging-based user's preference prediction mechanism by extending the widely known recommender algorithm, collaborative filtering (CF) in order to increase the user's satisfaction about the recommender service. The development of the prototype system shows the usefulness of proposed mechanism. And the experiment confirms that the proposed mechanisms improve the recommender system. © 2014 SERSC.