摘要
社会网络服务(SNS)用户的人脉关系研究大多采用图论的知识,对社会网络关系图的结点和边进行探讨,而没有考虑到用户自身的偏好。因此提出一种基于用户偏好的二级人脉推荐方法。利用最小均方误差(LMS)算法,把用户偏好合理地转化为用户偏好特征向量,用相似度度量方法来计算用户之间的相似度,以确定与用户偏好最相近的用户集,并完成用户的二级好友推荐。实验结果表明,该算法的好友推荐准确度较高。
The contacts relationship research of the customers in Social Network Service（SNS） is often based on graph theory.The nodes and the edges of the relationship graph of SNS are often discussed without considering the preference of customers themselves.Thus,a second-level contacts recommendation method based on preference of users is proposed in this paper.The LMS algorithm is used to transform the preference of customers into the eigenvectors of users preferences reasonably.We compute customers similarity by using similarity measurement method in order to decide the customer set closest to the preference of customers and to complete the recommendation of second-level friends of the customers.Experimental results show that this algorithm has higher accuracy in recommendation of good friends.
出处
《计算机应用与软件》
CSCD
北大核心
2012年第4期39-43,共5页
Computer Applications and Software
基金
国家自然科学基金项目(61172144)
关键词
社会网络
LMS算法
好友推荐
用户特征向量
Social network site LMS algorithm Recommendation of good friends Users eigenvector
