摘要
本文提出了一种结合人口分类特征计算用户相似度的协同过滤算法。实验结果表明,该算法可以改进相似用户的选取精度,从而提高推荐质量。最后,指出该算法可以解决网站刚启动用户数量不足时的“冷启动”问题。
A collaborative filter algorithm by taking demographic trait into consideration is proposed. Experimental results demonstrate that the algorithm can improve the accuracy to choose the similar user. At the end , it is pointed that the algorithm is applied in the condition that common rating items between the users are in sufficient.
出处
《安徽科技学院学报》
2007年第2期40-44,共5页
Journal of Anhui Science and Technology University
关键词
电子商务
推荐系统
协同过滤
基于内容的过滤
数据挖掘
E- commerce
Recommendation systems
Collaborative filtering
Content- based filtering
Data detecting
