基于关键词信息的微博用户行为度量分析研究  

Analysis and Measurement on the Behavior of Micro-blog Users Based on Keywords Information

作　　者：丁伟杰[1] 孔霆[2] Ding Weijie;Kong Ting(Department of Computer and Information Technology,Zhejiang Police College,Hangzhou 310053,China;Department of Experiment Center,Zhejiang Police College,Hangzhou 310053,China)

机构地区：[1]浙江警察学院计算机与信息技术系,杭州310053 [2]浙江警察学院实验中心,杭州310053

出　　处：《科技通报》2017年第5期124-128,133,共6页Bulletin of Science and Technology

基　　金：国家自然科学基金重点项目(U1509219;浙江省教育厅科研项目(Y201224395);浙江警察学院校级科研项目(20140622)

摘　　要：近年来,微博用户行为分析与挖掘逐步成为社交网络领域研究热点,本文基于微博传播机制及用户特点,提出了一种基于高频关键词信息的微博内容筛选算法,结合好友联合影响概率确定关键微博用户,并分析其影响力及具体范围。实验数据分析表明:本文设计的算法是可行的,与传统方法相比,本文算法在高频关键词提取和微博用户行为分析方面,能够有效提高话题传播中微博用户影响力度量的准确性。As the most prevalent social networking media,Microblog users’relationship mining attractsmore and more people’s attention.In order to excavate the influence of users’behavior information,wefirst preprocess the users by their characteristics on microblog,by which we eliminate the interference ofzombie powder,and advertisers,and thus draw a valid set of key users.Then by calculating theprobability of friends combined effect to determine whether the users are affected by a collection offriends.Further statistics the set of neighbor users will be affected by individual user.So we get theinfluence scope of the user’s behavior and the specific effected users.Experiments show that ourapproach is feasible and effective,and can accurately get the influence scope of the user’s behavior andthe specific effected users.

关 键 词：微博 用户行为 关键用户 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]