Adopting user profiles and behavior patterns in a Web-TV recommendation system

With recent advances in consumer electronics, it is noted that personalization and interactivity have become two challenging issues to further improve user experiences in watching TV. As there are thousands of TV programs broadcasted everyday, people may spend much time in finding their desirable programs. In view of this, we focus on developing a proper recommendation scheme in the Web-TV environment so that users can be recommended with appropriate programs. Specifically, we propose to estimate user preferences and ratings with both explicit and implicit information provided by users. Moreover, a prototype system which utilizes the adaptive recommendation approach is developed to illustrate the feasibility of the proposed scheme in this paper.