摘要
针对传统协同过滤推荐算法存在推荐精度低的问题,提出了一种基于社会网络的协同推荐方法。该方法融合了社会网络中用户的相似度与信任度,首先计算用户间的评分相似度;再由直接信任度与间接信任度加权得出用户信任度;最后综合用户相似度与信任度得出用户间的推荐权重,并以推荐权重来选取最近邻居集,为目标用户形成推荐。试验结果证明,该方法可有效提高推荐系统的推荐精度。
In consideration of the problem of lower recommendation precision in the traditional collaborative filtering recommendation algorithm,a new collaborative recommendation method is proposed based on social network.The similarities and credibility of users are integrated in the social network.Firstly,the similarities between the users are calculated based on the ratings,and then the credibility of users are calculated based on direct and indirect credibility.Finally,the similarities of user rating and the credibility of user'recommendation are integrated to get the weights of users' recommendations and get the nearest neighbor set and provide a more accurate recommendation.The experimental results show that the new method can improve the accuracy of recommendation.
出处
《长江大学学报（自科版）（上旬）》
CAS
2015年第6期29-33,4,共5页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI ＆ ENG
基金
教育部人文社科规划项目(13YJA630098)
安徽省高校省级自然科学研究项目(KJ2012B022)
关键词
社会网络
协同过滤
推荐精度
信任度
推荐权重
social network
collaborative filtering
recommendation accuracy
credibility
recommendation weight
