计算机工程与应用 ›› 2016, Vol. 52 ›› Issue (1): 55-60.

• 大数据与云计算 • 上一篇    下一篇

基于博文及网络结构信息的好友推荐方法

许超逸1,李德玉1,2,王素格1,2   

  1. 1.山西大学 计算机与信息技术学院,太原 030006
    2.山西大学 计算智能与中文信息处理教育部重点实验室,太原 030006
  • 出版日期:2016-01-01 发布日期:2015-12-30

Friend recommendation method based on micro-blogs and network structural information

XU Chaoyi1, LI Deyu1,2, WANG Suge1,2   

  1. 1.School of Computer and Information Technology, Shanxi University, Taiyuan 030006, China
    2.Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education, Shanxi University, Taiyuan 030006, China
  • Online:2016-01-01 Published:2015-12-30

摘要: 由于人们在书写用户属性信息时随意性和虚假性,使得在进行用户兴趣建模时用户属性无法得到有效利用。针对该问题,提出了一种基于兴趣偏好和网络结构的混合好友推荐方法。采用LDA主题模型对用户微博进行建模,从中挖掘用户兴趣,并依据同质性原理利用好友兴趣偏好对目标用户兴趣偏好进行修正。同时,采用一种新颖的基于网络结构的预测指标度量用户间的亲密程度。实验结果表明,与仅利用网络结构的推荐效果相比,加入用户兴趣后的模型在准确率及AUC指标上有显著提升,同时也可提高部分博文主题不明确用户的兴趣挖掘效果。

关键词: 好友推荐, 社交网络, 微博, 网络结构, 用户兴趣

Abstract: Due to the arbitrariness and falsity when people describe the attributes of users, it is invalid for us to apply the node attributes effectively while building the model of user interests. Aimed at this problem, a sort of hybrid friend recommendation method is proposed based on the interest preference and structural closures. At first, by using the LDA topic model, a model is constructed for users’ micro-blogs, so as to mining the interest of users. At the same time, the method refines the preference of the target users according to the principle of homogeneity through the interest preference of friends. Meanwhile, a novel prediction index based on network structure is proposed to measure the structure closeness between users. The experimental results indicate that the rate of accuracy and AUC has significant improvement considering the interest of user, compares with the effect of recommendation via only on the partial structure. The interest preference of friends has also to some extent enhanced the interest mining effect of part of users whose blog topics are not explicit.

Key words: friend recommendation, social network, micro-blog, network structure, user interest