（武汉大学信息管理学院 武汉 430072）

结合电子商务荐购系统的要素分析，构建能统一描述荐购系统输入信息、内部数据表示方式和各种荐购方法的用户、商品及交易关系模型。以此关系模型为基础，提出一种混合型荐购方法，并以此方法为核心构建能支持多种荐购方法自由切换、支持用户定制、支持信息推送服务以及商业排名服务的电子商务荐购模型。

(School of Information Management, Wuhan University, Wuhan 430072,China)

With the analysis of recommendation system essentials, this paper constructs a relation model of users, commodities,and transactions that could uniformly describe system input information,data representations, recommendation methods. Based on this relation model, proposes a hybrid recommendation method which serves as the core of our e-commerce recommendation model that supports free shifting between different recommendation methods, user customization、information pushing service and commercial ranking service.

G250

1J.Schafer, J.Konstan et al. E-commerce recommendation applications. Data Mining and Knowledge Discovery, 2001(5):115-153
2P.Resnick, H.Varian. Recommender systems. Communications of the ACM, 1997(3): 56-58
3C.Aggarwal, Yu,P.S. Mining associations with the collective strength approach. IEEE Transactions on Knowledge and Data Engineering, 2001(13):863-873
4M.Condliff, D.Lewis et al. Bayesian mixed-effects models for recommender systems. 1999.http://citeseer.ist.psu.edu/condliff99bayesian.html (Accessed Jan.5,2006)
5G.Adomavicius, A.Tuzhilin. Using data mining methods to build customer profiles. IEEE Computer, 2001(2):74-82