摘要
协同过滤算法作为一种成功的个性化推荐技术已经被应用到很多领域中。传统的协同过滤算法中用户相似性的计算只考虑了用户评分信息而没有考虑到用户的社会背景信息,针对这个问题,本文提出了基于用户社会信息的相似度计算方法,实验表明,改进后的协同过滤算法能更好地反映用户兴趣,提高推荐精度,在推荐效果方面得到了更好的改善。
Collaborative filtering is one successful personalized recommendation technology, and is extensively used in many fields. Collaborative filtering algorithms only consider users＇ mark information. They do not consider users＇ social information. To solve this problem,the paper puts forward the calculation of social information similarity. Experimental results show that our proposed algorithm outperforms traditional collaborative filtering algorithm.
出处
《桂林电子科技大学学报》
2009年第3期234-237,共4页
Journal of Guilin University of Electronic Technology
关键词
个性化服务
协同过滤
社会信息
相似度
personalized service
collaborative filtering
social information
similarity
