Key points are not available for this paper at this time.
The abundance of DTV (digital television) programs precipitates a need for new tools to help people personalize interesting TV content. We developed an adaptive assistant: TV3P (TV program personalization for PDR), which observes users' viewing behaviors in the background, updates users' profiles continuously arid autonomously, and then filters and recommends programs for different users according to their respective preference information. The novel aspect of this system is the evaluation of how much time and effort it takes the system to learn new preferences once it already is biased by old preferences. This has not been proposed in any other recommender systems before. It was also proved to match real world users whose preferences can change over time. Another attractive aspect of TV3P is it's employing an implicit and explicit profiling scheme.
Yu et al. (Sun,) studied this question.