基于社交网络大规模行为数据的用户关系研究  

ON USERS RELATIONSHIP BASED ON LARGE-SCALE BEHAVIOUR DATA IN SOCIAL NETWORKS

作　　者：刘晶[1,2] 李琳[3] 李石君[1] 

机构地区：[1]武汉大学计算机学院,湖北武汉430072 [2]中南民族大学计算机科学学院,湖北武汉430074 [3]武汉数字工程研究所,湖北武汉430074

出　　处：《计算机应用与软件》2016年第7期38-41,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(61272109);中央高校基本科研业务费专项资金项目(CZY15006)

摘　　要：用户关系是构成微博社会网络的基础。用户关系的分析可以帮助更好地研究社会网络的构成、消息传播模式等多个方面。对超过百万用户的海量微博数据进行分析处理,利用信息论理论分析比较用户微博行为的特点,构建用户活跃交互网络并观察交互网络的动态性,分析社交网络用户群体的在线行为模式及特点。实验表明在微博的交互活动中,用户的直接交互关系相对稳定,不因时间的变化而变化,而用户的转发对象会不断地变化,即用户实际关注的群体是动态变化的。User relationship is the basis of microblogging social network formation. To analyse users relationship can help the better study in regard to the formation of social networks and the messages dissemination patterns,etc. In this paper we analyse and process massive microblogging data of more than one million users,and use information theory to analyse and compare the features of users microblogging behaviour,construct active users interaction network and observe its dynamics property,as well as analyse the online behaviour patterns and features of user groups in social networks. Experiments show that in microblogging interactions,direct interactive relationship between users are relatively stable and will not change along with the time going,while their forwarding objects are constantly change,that is,the groups actually concerned by the users are dynamically changing.

关 键 词：社会网络 用户行为 微博 交互 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]