Research And Implementation Of E-commerce Recommender System Based On Collaborative Filtering Technology

With the rapid development of the Internet, online shopping has gradually become one of the consumer behavior. However, as the commodities explode in number and size, customers cannot often find his/her satisfied products in time when facing mass data. In order to make customer find his/her needs easily and fast, it is particularly important to provide real-time personalized recommender on the e-commerce sites, this can not only increase the sales of goods, but also enhance the loyalty of customer to the sites.This paper starts with the introduction of basic situation of the e-commerce personalized recommender system, expounds the e-commerce recommender system framework, recommender process, structure and classification, etc. Then, this paper probes into the personalized recommender technology, introduces the concept of collaborative filtering technology in detail, and focuses on the analysis of the collaborative filtering technology, the principle of the process and the calculation method of the similarity about the technology. As the same time, the advantages and the deficiency of collaborative filtering technology is described by the paper.The paper thoroughly studies the methods about how to improve the quality of recommender, introduces the common methods to solve the sparse and the cold-start problems according to the deficiency of collaborative filtering technology. Then, the paper puts forward its own research scheme and makes a detailed analysis about how to relieve the two problems effectively. Meanwhile, the paper proves the feasibility of the scheme using the related data set.On the basis of the thoroughly study of recommender system which bases on collaborative filtering technology, this paper constructs a recommender engine which bases on collaborative filtering technology, expounds the process and method of construction and tests the feasibility using a data set. Finally, this paper apply the engine to a recommender system about the e-commerce film, the system realizes the online personalized recommender in real time.