面向话题的微博网络测量研究

发布时间：2018-02-27 12:16

本文关键词： 内容特征网络结构用户行为小世界特性局部波动率影响力分析　出处：《通信学报》2013年11期 　论文类型：期刊论文

【摘要】：针对话题生成网络的动态时序特性,设计定量计算方法,从微博内容、网络结构、用户行为角度开展面向话题的新浪微博网络测量研究,结果发现:少数微博被大量转发,转发次数与对应微博数呈现近似的幂率分布;话题热度呈现明显的突发性和变化趋势,局部波动率能够有效地在大量背景微博中发现突发话题;基于话题生成的转发网络的小世界特性并不明显,且密集的关注关系不一定引发频繁的转发行为;传播能力强的话题中含有较大比例的持续参与用户,用户行为的话题相关性能够有效检测潜在关键用户。测量结果有助于了解话题生成网络的内容传播特点、网络结构特性及用户行为模式,测量指标能够有效应用于微博话题影响力分析等相关研究。
[Abstract]:In view of the dynamic temporal characteristics of topic generation network, a quantitative calculation method is designed to carry out the research on topic oriented Sina Weibo network measurement from the aspects of Weibo content, network structure and user behavior. The results show that a small number of Weibo has been forwarded in large numbers. The frequency of forwarding is similar to the corresponding Weibo number, the topic heat shows obvious sudden and changing trend, and the local fluctuation rate can effectively find the burst topic in a large number of background Weibo. The small-world characteristics of topic-generated forwarding networks are not obvious, and the intensive attention relationship does not necessarily lead to frequent forwarding behavior; the topic with strong communication ability contains a large proportion of continuous users. The topic correlation of user behavior can effectively detect potential key users. The measurement results are helpful to understand the characteristics of content dissemination, network structure and user behavior patterns of topic generation network. The measurement index can be effectively applied to Weibo topic influence analysis and other relevant research.
【作者单位】：中国科学院计算技术研究所;中国科学院大学;国家计算机网络应急技术处理协调中心;
【基金】：国家自然科学基金资助项目(61170230) 国家科技支撑计划基金资助项目(2012BAH46B01)~~
【分类号】：TP393.09

【参考文献】