The Research And Implementation Of Movie Recommender System Based On Hybrid Recommendation

With the development of people’s entertainment, the spiritual entertainment is particularly important. The film as a universal entertainment has been paid more and more attention. However, how to find their favorite film in the massive movies has become a problem. Now personalized recommendation algorithm has come into being, it is to solve how to find the critical information, and how to recommend information to the users to meet their requirements. For a recommender system, how to use the database effectively and rationally is a focus, and how to use the historical information and users’behavior to predict users’potentially interest is the focus of the study of the system.The main content of this paper is to help users to extricate themselves from the massive movie information. Avoiding letting users be interfered with too much information, the system can automatically recommend the interested movies to the users. This paper describes several recommendation algorithms. Each algorithm has its own advantages and disadvantages. Using one kind of algorithms cannot complete the actual demand. Therefore, in practical application, mixing two or more algorithms together to recommend is particularly important. Introducing each recommendation algorithm is the foundation of this paper. Mixing each recommendation algorithm is the key point of this paper. The purpose of this paper is to understand the accuracy and efficiency of hybrid recommendation algorithm in practical application.The core of the hybrid movie recommendation system is three-tier architecture based on B/S. The three-tier architecture has a presentation layer, a business layer and a data layer. The system achieved in this file combines this kind of three-tier architecture, and uses UML modeling language to design.First, research the principles and characteristics of the recommender system, and then build theoretical basis of the movie recommender system. There are several basic recommendation algorithms. They are collaborative filtering algorithm, content-based recommendation algorithm and network-based recommendation algorithm. We need understand the advantages and disadvantages of each algorithm, in order to use the reasonable algorithm during the course of implement of the system.Second, research the business needs of movie recommender system after having discussed how to achieve system and what kind of technology is needed. We need module design, the overall framework design, and the recommendation algorithm design. The database of the system is built on this basis of needs analysis.Finally, after implementation of the system, the hybrid movie recommender system can be used for users. And users can use this system to manage movie information and to find their potentially interested movies.