摘要
微博的普及导致微博平台数据量日益增长,因此从海量微博中快速准确地为微博用户推荐好友成为了巨大挑战。用户的社交网络和微博文本在一定程度上体现了用户的价值观和兴趣爱好,有相似兴趣的微博用户更有可能成为朋友。基于上述事实,以用户微博文本相似度为似然函数,使用K-means聚类对微博用户聚类,得到微博用户社交圈;在社交圈内部迭代计算用户之间的相似度,同时计算用户对其所在社交圈中其余用户的信任度;最后,根据用户之间的相似度和信任度完成微博好友推荐。实验结果表明,该算法优于传统的基于社交网络拓扑图的好友推荐方法。
The popularity of micro-blog has caused an increasing amount of data on micro-blog platforms.Therefore,recommending friends quickly and accurately for micro-blog users has become a great challenge from the massive micro-blog.The users′micro-blog behavior and social network largely reflect the users′values and interests.Micro-blog users who have similar interests are more likely to be friends.In view of the above facts,the topic similarity of user micro-blog text is used as a likelihood function,and K-means clustering is used to cluster the micro-blog users,then the micro-blog users′social circles are obtained.In social circles,iterative computation of the similarity between users is done,while trustworthiness of the users to the rest of the users in the social circle.Finally,according to the similarity and trust between users,the recommendation of micro-blog friends is completed.The experimental results show that the algorithm proposed in this paper is superior to the traditional recommendation method based on social network topology.
作者
朱明玮
唐莫鸣
ZHU Ming-wei;TANG Mo-ming(School of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650504,China)
出处
《软件导刊》
2018年第6期62-66,共5页
Software Guide
关键词
社交圈
信任度
朋友推荐
微博
social circle
trust degree
friends recommended
Micro-blog
作者简介
朱明玮（1992-），女，昆明理工大学信息工程与自动化学院硕士研究生，研究方向为自然语言处理；;唐莫明（1993-），男，昆明理工大学信息工程与自动化学院硕士研究生，研究方向为自然语言处理和信息检索。
