Abstract:

As an important social media, many scholars have studied the influence of Weibo users, but most of the influence evaluation algorithms are based on the static attributes of users in Weibo topics or the behavioral characteristics of users after the occurrence of Weibo topics. Through the comprehensive analysis of repost, comment, like and combining with emergent computation model, this paper proposes a user influence sorting algorithm based on Swarm model. SMRank algorithm can calculate the influence of users in each time period during the occurrence of Weibo topics, which provides a new method to calculate the influence of users on Weibo topics. Based on the real SinaWeibo data, the results show that the SMRank algorithm can effectively find higher influence users and calculate the users’ influence at different moments.

Key words: emergent computation, Swarm model, user influence

摘要：

微博作为一种重要的社交媒体，许多学者都对微博中用户的影响力进行研究，但大多数影响力的评价算法都是根据微博话题中用户的静态属性或微博话题发生后用户的行为特征对用户影响力进行评价。从用户的转发、评论和点赞三种行为入手，结合突现计算模型，提出一种基于Swarm模型的用户影响力排序算法，SMRank算法可以在微博话题发生的过程中对用户每个时间段的影响力进行计算，给出了一种计算微博话题用户影响力的新方法。通过使用真实的微博话题数据进行实验，结果表明提出的SMRank算法可以有效地发现微博话题中影响等级较大的用户，并能计算出不同用户不同时刻的影响力。

关键词: 突现计算, Swarm模型, 用户影响力