首页 | 本学科首页   官方微博 | 高级检索  
     检索      

种多特征微博僵尸粉检测方法与实现
引用本文:王越,张剑金,刘芳芳.种多特征微博僵尸粉检测方法与实现[J].中国科技论文在线,2014(1):81-86.
作者姓名:王越  张剑金  刘芳芳
作者单位:重庆理工大学计算机科学与工程系,重庆400054
基金项目:重庆理工大学研究生创新基金资助项目(YCX2012317)
摘    要:微博中僵尸粉的大量出现,不仅对微博影响力计算与社交网络关系分析带来了新的挑战,而且对用户带来了社交诚信危机。首先对微博僵尸粉进行概念上的定义;其次通过用户个人信息、用户微博内容和用户链接关系分析僵尸粉与普通用户之间的不同特征,并训练了一个基于C4.5决策树的僵尸粉分类系统;最后使用新浪微博数据对系统进行评估,结果显示该系统对微博僵尸粉有92.8%的判别准确率与92.8%的召回率。

关 键 词:微博  僵尸粉  虚假用户  新浪

Detection of micro-blog zombie fans based on multi-features
Wang Yue,Zhang Jianjin,Liu Fangfang.Detection of micro-blog zombie fans based on multi-features[J].Sciencepaper Online,2014(1):81-86.
Authors:Wang Yue  Zhang Jianjin  Liu Fangfang
Institution:(Department of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China)
Abstract:The large number of micro-blog zombie fansnot only brings new challenges to the computation of micro-blog influence and the analysis of social networking relationship,but also makes users fall deep into a social credit crisis.In this paper,we first-ly define the concept of micro-blog zombie fans,then we analyze the differenceof characteristics between zombie fans and ordinary usersthrough users’personal information,users’micro-blogs content and the users’linking relationship,and a zombie fans clas-sification system based on C4.5 decision tree algorithm is trained;finally we use the Sina data to evaluate the system,and the re-sults show that the system has an accuracy rate of 92.8%and a recall rate of 92.8%on discrimination of micro-blog zombie fans.
Keywords:micro-blogs  zombie fans  fake user  Sina
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号