Design And Implementation Of A Collaborative Filtering Personalization Recommender System Based On Slope One Algorithm

With the popularization of Internet, commercial websites flourish. How to improve the effectiveness of commercial Web sites, especially the use of personalized recommender technology to provide personalized service to realize has gradually become a hot spot subject which can arouse interest.Commercial sites'development from the"Web site centered design"to"user centered design " will be inevitable tendency. However, commodities recommended mode in today's most of commercial sites: hot commodities; related commodities; information history the user browsed. Because the first two recommended mode has not taken personalization features into account, they do not have any personal characteristics. The third recommended mode is personalize in a degree but on the majority of sites they also just remains on a stage of building a information database only based on the user's browsing history, and they didn't make horizontal information integration, So it is valueless for collaborative recommendation.This paper have a discussion on the theory about recommender system base on collaborative filtering technology, and the use in engineering applications, and deficiencies and difficulties of this system. Further, this paper research on the Slope One algorithms which is built by Prof. Daniel Lemire in 2005 algorithm is a kind of Item-Based recommender algorithms .It can be applied to all kinds of websites through the Internet-based business, such as online store, and websites they provide"intangible" products, such as articles, news, music, movies. It can be used to help Web sites to provide personalized recommendations, to tap potential customers and old customers' potential buying power.This paper used J2EE Framework to implement the system, with emphasis on personalized recommender module implementation. Firstly, it uses the popular framework of "JSF + Spring + Hibernate" as the system architecture, and their engineering experience on the operational level to layer the system. Secondly, it use current achievements and problems-oriented personalized recommender system based on personalized recommender system as primary expressions , personalized recommendation system based on Slope One algorithm has designed and implemented. To an extent I also give assessment and I make some exploring on algorithm in the application of business systems.