中文摘要：

微博是网络舆论的重要聚集地,在突发事件危机应对中发挥着巨大的作用。以新浪微博为基础,首先结合标签传播算法、随机扩散模型、内容分析法等技术,提出了突发事件中用户动态关系的社群发现及舆情传播特征分析的研究思路,然后以实际突发事件作为研究对象进行了实证研究,最后发现结合用户链接相似度的标签传播算法能较好地发现微博中的用户社群;用户转播率可较好描述社群中舆情传播程度,并以此区分不同社群的信息传播能力;不同社群的微博内容差异较大。因此重点性地识别有影响的社群有利于危机应对。

英文摘要：

Microblog become a public opinion gathering place and plays an important role in crisis response. Firstly, based on the Sina microblog the paper combined with PLA, stochastic diffusion model and content analysis, putting forward a method to find user communities and a way to analysis emergency propagation characteristics. Then we make an empirical research with actual data. Finally we find that the PLA model combining user link similarity is effective in micro blog community discovery; the characteristic of communities is different from each other because of the differences of structure and users＇ relationships; the user node retransmission rate can describe the degree of public opinion communication and the in-formation transmission ability; users＇ behavior are not the same at all, identifying more influential community is necessary to crisis resolve.