抄録
I-018
Latent factor model for user preferences and prediction of TV program in IPTV environment
◎Quoc Hung Dinh・Pao Sriprasertsuk・亀山　渉（早大）
Many movie and TV program recommendation systems are available both in markets and research institutes, nevertheless those systems are not exploited deep enough to find the user preferences and user biases to predict effectively user evaluation of the recommended TV programs. Thus, our work aims to explore a way to integrate user viewing history, user evaluation and user bias in TV programs to build the recommendation system in IPTV in which user will potentially watch them with the highest possibly assessment. The proposed system is implemented by the approach of latent factor model.