Design And Implementation Of Recommendation System Based On Collaborative Filtering

The rapid development of information technology is to make people gradually entered the era of information overload, the generation of recommendation system is one of the effective methods to solve the problem. And the collaborative filtering recommendation system is commonly used in the method, its biggest advantage is that don’t have to consider the content of the recommended items, just according to the similarity of making social recommendation to target users. Although collaborative filtering has been widely used in all areas of life, but it is still there is a cold start,sparse, accuracy and scalability.This article first do to recommendation systems are an integral part of the introduction, research and analyses some common recommendation technology and the collaborative filtering algorithm, designs and realizes a movie recommendation system based on B/S structure. This paper presents a new similarity calculation method, combined with the similarity of user interest and active users and hot items influence on personalized recommendation, improve the accuracy of the similarity calculation. Similarity filter parameters is proposed in this paper, the selection of nearest neighbors collection was further optimized, while the algorithm score prediction module introduces the evaluation factors and control factors on user-based and item-based prediction scores were weighted mixed. The final design and implements a hybrid collaborative filtering algorithm based on user- item. With the collaborative filtering method based on the user and based on the item on the selected data set to do the experiment contrast. Experiments show that the hybrid algorithm can ease the data sparseness to a certain extent and can greatly improve the accuracy of the recommendation.Based on this hybrid algorithm, this paper implements a movie recommendation system. From the functional requirements, module partition, first has carried on the overall system architecture of system design. Then in detail using the J2 EE three-tier architecture thought and for Spring MVC, My Batis framework to build the whole process of design and implementation of the system. System can be very good to provide users with personalized movie recommendations, further shows the effectiveness of the proposed hybrid algorithm.