Design And Implementation Of The Movie Recommendation System Based On Matrix Completeness

In the information age,the explosive growth of network data brings not only better service and more convenient experience,but also information overload.As far as movie websites are concerned,the growing number of movie data resources and movie viewers make it impossible to select views' favorite movies in a short time.So how to recommend accurately for viewers has become a hot issue in the research and design of film system.From then on,this paper proposes a recommendation algorithm based on matrix completeness to design the movie system,and implements this assumption.The specific research contents are as follows:(1)Designing algorithms to alleviate the problems of sparsity,scalability and similarity distortion in collaborative filtering.The core idea is to fill in the scoring matrix with SVT algorithm based on matrix completeness according to the low rank of the scoring matrix;then divide the users by mini batch kmeans clustering method to build the nearest neighbor relationship between users,so as to reduce the search time of users when building the nearest neighbor;finally,recommend top-N with similarity algorithm with time factor.(2)Solving the cold start problem of the system,on the one hand,use the registration information of new users to calculate the feature preferences of new users as the recommendation standard;on the other hand,use the category and label information of movies to calculate the movie similarity to get the recommendation information.(3)Conducting a comprehensive requirements analysis of the system,and proposes and constructs the overall framework of the system according the requirements analysis,and designs each functional module in details.(4)According to the design of each functional module,each module is implemented,and the system is tested from the perspective of function and performance after the system is implemented.Therefore,this paper designs and implements a movie recommendation system based on Django.The test results show that the system can solve many problems,such as sparsity,cold start and so on,and meet the multi-faceted requirements of system function and performance.