Abstract: The rapid development of e-commerce has promoted the growth of online auction business based on C2C context. However, the ever-increasing user size and auctioned goods cause the problem of information overload, and how to enhance the customer loyalty becomes a critical issue faced by most online auction websites. One way to overcome the problem is to use recommender systems to provide personalized information services. Since there exists much difference between B2C and C2C context, it is a new challenge for usere to apply recommender systems to the latter setting. This paper analyzes the user behaviors on the auction website and constructs the user preference model under the C2C context. Then the collaborative filtering technique is used to recommend auctioned goods.

Key words: C2C, Recommender systems, Collaborative filtering, User modeling

摘要： 随着电子商务的发展，基于C2C环境的在线拍卖也迅速发展起来，用户数和拍卖商品数的急剧增加，使得信息过载和如何提高客户忠诚度这一问题凸现出来。而推荐系统则成了解决这一问题的手段之一。但是C2C和B2C环境存在很大的不同，对推荐系统的应用提出了一定的挑战。该文对用户在拍卖网站的行为进行了分析，在此基础上建立了用户的偏好模型，利用协同过滤技术进行拍卖商品的推荐。

关键词: 顾客对顾客, 推荐系统, 协同过滤, 用户偏好模型