Research And Application Of A Collaborative Filtering Algorithm Based On Trusted Similarity Propagation

With the continued development of the Internet, e-commerce in recent years entered a period of rapid growth. More and more product information generated in e-commerce makes it difficult for users to quickly find their favorite products. At the same time, e-commerce companies are faced with the problem that how to provide users with interesting products in time to improve sales. Recommendation system is an effective way to solve such "Information Overload" problem and Collaborative Filtering is one of the best and widely used technologies in recommendation Systems. But collaborative filtering is also facing data sparsity, cold start, scalability and other issues.Based on the comprehensive analysis of e-commerce recommendation system and several common recommendation technologies' advantages and disadvantages, we propose a new collaborative filtering algorithm based on trusted similarity propagation for the data sparsity problem of collaborative filtering. First we introduce a way to build user trust model in the traditional user-based collaborative filtering algorithm, and then applied this trust relationship to the processes of similarity propagation and rating prediction. Experiments show that in the case of sparse rating data, the algorithm can improve the coverage and accuracy of the predicted rating.Meanwhile, we use this algorithm in a recommendation module of an e-commerce site. With the function of product recommendation, we improve customers' satisfaction.