Personalized TV program recommendation based on TV-anytime metadata

The expansion of TV channels in digital era is resulting in a soaring number of program contents available to viewers. The advent of Personal Digital Recorder makes users face more viewing selections. Besides watching a program that is being broadcast, users can select something favorite to record so that they can watch these contents whenever they like. Viewers are really exposed to TV program overload. Consequently, the personalized TV recommendation is necessary to automatically customize program listings. This paper proposes a personalized TV system based on TV-Any time metadata model. In addition to the traditional functions, such as program navigation and search, this system is characterized by recommending program with high preferences that are automatically generated on the basis of usage history.