User's Interests-based Movie Recommendation in Heterogeneous Network

For large data set of movies and various users' interests, recommender systems should pay attention to time consumption and personal interest. However, existing information filtering techniques rarely research users' interests in movies. In order to implement personalized movie recommendation effectively, we propose user's interests-based movie recommendation in heterogeneous network (IMRHN). First, we divide the heterogeneous network consisted of movies, movie genres and users into several sub-networks based on movie genres, and then mine active users' interest probability on every movie genre through other users' influence. At last, we recommend higher marked movies for each certain eligible movie genre that attracts more active user's interest. Experimental analysis suggests that IMRHN can achieve personalized recommendation and perform better in time consumption than other several algorithms.