面向舆情主题的微博用户行为聚类实证分析

Empirical Analysis of Micro-blogger Behavior Clustering on Public Opinion Topics

网络舆情管理是社会管理创新的重要内容,面向舆情主题的用户行为分析是网络舆情管理的热点问题。对三个不同舆情主题下的微博用户行为进行了聚类分析,发现信息传播中的微博用户基本可分为"一般关注"型、"主动参与"型、"信息传播"型三类。"一般关注"型用户通常人数众多,他们多为舆情信息的接受者。"主动参与"型用户类中用户人数较少,他们大多以自己为中心,向朋友发布舆情信息,这类用户在不同主题下其行为特征差异很大;"信息传播"型用户类中人数很少,是舆情领袖或重要信息转发者。结果表明可以尝试将监测"信息传播"型网民和引导"主动参与"型网民结合起来,为大数据下微博舆情信息管理提供智力支持。

The management of network public opinion is an important practice of innovative social administration, whereas analysis of mi-cro-bloggers behaviors with different public opinion topics is an issue of great concern. Under three different public opinion topics, this paper clusters micro-bloggers into three groups according to their behaviors in the process of information propagation： observers, active participants and propagators. Observers are obviously dominant in number and are usually information recipients. Active participants, small in quantity, are often self-centered and prone to disseminate information to their friends, and their behaviors characteristics may dif-fer on different public opinion topics. In contrast, propagators, few in quantity, play the role of opinion leader or are critical dissemina-tors. The findings imply that a combination of both monitoring propagators and guiding active participants could be an effective strategy for micro-blog public opinion management in the era of Big Data.