Hyper User Preference Access in Personalized Recommendation System
个性化推荐系统中的混合用户偏好获取

For the sparse data of user preferences in personalized recommendation system, a new hybrid user preference access is presented. The accurate foundation, considers the user residence time, mouse clicks, and page scrolling time, receiving implicict ratings. The user preference matrix is constructed with implicit and explicit rating, providing a data base for a recommended algorithm. Experiments proves that the hyper user preferences access is feasible and effective.