• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2013, Vol. 35 ›› Issue (3): 140-144.

• 论文 • 上一篇    下一篇

微博中基于用户偏好的信息传播研究  

王金辉,贺利坚,张伟,童向荣   

  1. (烟台大学计算机科学与技术学院智能信息处理实验室,山东 烟台 264005)
  • 收稿日期:2012-05-09 修回日期:2012-09-21 出版日期:2013-03-25 发布日期:2013-03-25
  • 基金资助:

    国家自然科学基金资助项目(61170224);山东省自然科学基金资助项目(ZR2011FL018,ZR2012FL07);山东科技发展计划资助项目(2012GGB01017);山东省教育厅科技计划资助项目(J11LG35,J10LG27)

Research of user preference based
information dissemination in microblog  

 WANG Jinhui,HE Lijian,ZHANG Wei,TONG Xiangrong     

  1. (Laboratory of Intelligence Information Processing,
    School of Computer Science and Technology,Yantai University,Yantai 264005,China)
  • Received:2012-05-09 Revised:2012-09-21 Online:2013-03-25 Published:2013-03-25

摘要:

对用户行为的分析是Web社会网络分析中的重要课题之一,包括对用户个体和群体的行为和偏好进行建模。为预测Web社会网络中的信息传播趋势,以微博为平台,基于对用户历史行为信息的统计分析,抽取出影响用户决策的因素,进而定性地分析各因素之间的关联关系,从而挖掘出用户潜在的偏好。在此基础上,利用CPnets偏好表达工具建立了用户偏好模型,直观有效地表达各因素之间的偏好关系,同时可根据其导出图输出最优的特性来选取最佳特征项组合,以便用户进行转发决策度量。

关键词: 社会网络分析, 偏好建模, 微博, 决策

Abstract:

Analyzing users’ behaviors is one of the important topics about web social analysis. It includes modeling the behavior and preferences of individuals and groups. For predicting the trends of information diffusion in web social networks, with Microblog for the platform, varies of factors to let the users make decisions are extracted, the relationships between these factors and information processing are analyzed and some potential preferences are mined basing on the statistics analysis of users' history behavior information by the researchers. Therefore, the users' preference model is established by CPnets, which could represent the preferences among all the attributes,and according to its induced graph the best combination of features will be selected to measure the possibility of forwarding a topic by one person.

Key words: social networks analysis;preferences modeling;microblog;make decisions