中文摘要：

基于社会网络分析方法（SNA）研究事故灾难舆情传播的网络结构、节点地位对信息的传播效率的影响等问题，力图提出针对关键节点的网络舆情治理策略。以最近的四件重大事故灾难为例，运用Gephi分析信息传播网络拓扑结构，探究事故灾难舆情在网络结构中的相似性，以排除网络结构对舆情传播的影响，进而基于SNA筛选出舆情中的两类关键用户：关键信息发布者、关键事件关注者，并运用武汉大学社会计算平台ROSTCM6分析关键节点及全体网民的情绪演化进程，提出以关键节点为中心的事故灾难舆情治理策略：关键信息发布者ID可作为事故灾难舆情案例库的一项构成成分，进而将舆情治理工作提前到舆情爆发之前，同时，通过其发文内容的情绪控制能够有效引导公众情绪；利用关键事件关注者的情绪预测作用，能够提高网络舆情预警机制的准确度。

英文摘要：

This paper aims to explore the location of nodes and the network structure＇s mtluence on acc-aent puonc opinion dissemination, trying to put up forward the strategies of public opinion which aims at the key nodes. Based on the four recent significant accident, Gephi is explored to generate the topology of public opinion spread network in the events, exploring the similarity of accident public opinion network structure, excluding the influence that network structure produced to public opinion dissemination so that two kinds of key users can be selected based on SNA： key event follower and key message sender. ROSTCM6 is explored to analyze the emotion of the public and key nodes, offering new methods for network public opinion management which takes the key point as center： the key message sender＇s ID can be used as one constituent of accident public opinion case base to put forward the dealing work be- fore the explosion, and at the mean time public＇s emotion can be leaded by what inside the passage efficiently. Im- prove the accuracy of public opinion monitoring early-warning mechanism by making full use of key event followers＇ emotion predictive effect.