信息网络安全 ›› 2015, Vol. 15 ›› Issue (6): 73-78.doi: 10.3969/j.issn.1671-1122.2015.06.012

• 技术研究 • 上一篇    下一篇

基于PageRank和用户行为的微博用户影响力评估

张俊豪, 顾益军(), 张士豪   

  1. 中国人民公安大学网络安全保卫学院,北京102623
  • 收稿日期:2015-03-09 出版日期:2015-06-20 发布日期:2018-07-16
  • 作者简介:

    作者简介: 张俊豪(1990-),男,河南,硕士研究生,主要研究方向:网络安全与数据挖掘;顾益军(1968-),男,江苏,副教授,博士,主要研究方向:网络情报技术;张士豪(1992-),男,山西,硕士研究生,主要研究方向:网络安全与数据挖掘。

  • 基金资助:
    公安部重点研究计划项目[2011ZDYJGADX016]

The Microblogging User Influence Assessment Based on PageRank and User Behavior

ZHANG Jun-hao, GU Yi-jun(), ZHANG Shi-hao   

  1. School of Cybersecurity, People’s Public Security University of China, Beijing 102623, China
  • Received:2015-03-09 Online:2015-06-20 Published:2018-07-16

摘要:

微博用户在消息传播过程中起到一个至关重要的作用,具有强大影响的微博用户是舆论形成、传播引导的关键因素。为了更准确地评估微博用户的影响力,文章提出了基于PageRank和用户行为的用户影响力评估(UIA,User Influence Assessment)算法,该算法对用户自身的活跃度和用户之间的联系度进行综合考量,通过对用户的行为分析,可以提取并量化出影响用户自身活跃度和影响用户之间联系度的相关因素,进而计算出权值分配比例。该算法既避免了PageRank的主题漂移现象,又系统化地衡量了用户影响力,具有较高的准确性。

关键词: 用户影响力评估, PageRank, 用户行为, 活跃度, 联系度

Abstract:

Microblogging users play a vital role in the process of news propagation, and Microblogging users with strong influence are the key factor in public opinion formation, spread and guide. In order to more accurately assess the influence of microblogging users, the UIA (User Influence Assess) algorithm, this paper presented based on user behavior and PageRank, which assess the user's own activity level and the connection degree among the users integratively, present and quantify the related factors that affect the user's own activity level and the connection degree among the users based on that we can analyze user’s behavior, and then come to the proportion of weighting. This algorithm avoided the topic drift phenomenon of the PageRank algorithm, and also systematically measured the user’s influence with highly accuracy.

Key words: UIA, PageRank, user behavior, activity degree, contact degree

中图分类号: