基于用户社区的商品推荐方法

Method for Commodity Recommendation Based on User Community

提出一种以用户社区服务系统为基础,面向社区新用户的商品推荐方法。根据现有用户的历史行为对用户进行社区划分,得到社区划分的结果模型,对于一个新来的用户运用这个模型将其归入相应的社区中,再根据这个社区的特征有目的地为新用户进行商品推荐。文中对该方法所涉及的基于信息熵的社区发现算法以及基于网络社区的协同推荐算法等关键问题的实现思路进行了详细阐述。

In this paper a new method for commodity recommendation is presented, which is based on user community service system. According to the history of the existing user behavior to divise the user community, the division of the community for the model, use this model divise new user into the corresponding community, and then according to the characteristics of the community recommended commodity for new users. The solutions for the key pmblemsin this method are given in details, including algorithm of community detection based on information entropy and collaborative recommendation algorithm based on community.