Research And Implementation Of Personalized Movie Recommendation System

With the development of Internet and the rapid popularization of mobile terminals,the quality of people’s lives has been greatly improved. The number of movie provided for users to watch on network is quite large and numerous in variety, however it takes long time for users to choose movies which they are interested in and the utilization of movie source is too low. Personalization recommender technique can discover the valuable information from mass information and provides users with personalized service, so it can solve the question of low utilization of movie source.In a personalization recommender system, the quality of recommender algorithm has a direct impact on the quality of recommender, so recommender algorithm has been a focus and hotpot of research for a long period. Though the application of collaborative filtering algorithm is quite widely, the recommender quality of user-based collaborative filtering algorithm is not very high. There are two reasons for the low quality: first, the user’s features are not successfully used in the algorithm; second, there is data sparsity problem of the rating matrix which stores users’ rate on item. On the above question, a new collaborative filtering algorithm based on users’ comprehensive similarity was proposed which introduces the users’ feature into collaborative algorithm. The comprehensive similarity is weighted summation of feature similarity and rating similarity weighting and user the theory of propagation of similarity to reduce the influence of data sparsity. At last, design experiments with standard data set and list the experiment results and do a detail analysis on the experiment results. The experiment results show that collaborative filtering algorithm based on users’ comprehensive similarity can improve the accuracy of recommendation.On the basis of research on recommender algorithm, a movie recommender system was designed and implemented. The movie recommender system is three-tier MVC architecture based on b/s model. First of all, do detail demand analysis of movie recommender system. Second, on the basis of the needs analysis do the overall framework of design, functional module design, recommender algorithm design and database design and list the technology roadmap and development environment. Next,develop and implement the movie recommender system and show the realization of the system’s key technology, including Personal Information Management, Movie Information Management and Recommender engine. Last, show the running status anddo a functional test and performance test on the system.