摘要
针对目前协同过滤算法浪费了大量用户信息的情况,本文提出了一种考虑用户背景信息的协同过滤算法。该方法提出两个用户模型,通过计算两个用户模型的相似度找出目标用户的伪邻居,最后通过评分矩阵计算相似度,找到目标用户真正邻居并且做出推荐。实验证明,该方法能更合理、准确的为目标用户做出推荐。
Currently,Collaborative filtering Algorithms ignore lots of information of users,in view of this situation,a method that sufficiently utilizes users' background information is proposed. This method proposes two models:user basic model and user preferences model. After computed similarity of two model respectively,we can find the pseudo neighborhood of the target user. Target user finds the genuine neighborhood by computing similarity with pseudo neighborhood.
出处
《微计算机信息》
2010年第36期197-198,159,共3页
Control & Automation
关键词
推荐系统
协同过滤算法
相似度计算
recommendation system
collaborative filtering algorithm
similarity computation