情感与主题协同分析的突发事件舆情图谱构建

Construction of Emergency Public Opinion Map Based on Collaborative Analysis of Emotion and Theme

分析网民情感在重大突发事件中不同舆情周期和不同主题的分布情况,构建基于情感与主题动态协同分析的重大突发事件舆情图谱,可为重大突发事件舆情引导提供帮助。在对灾难事件情感词典构建方法、LDA主题文本挖掘方法和舆情图谱构建方法进行分析的基础上,采集并处理新浪微博相关热搜话题数据,进行基于网民情感的舆情周期和主题挖掘及协同分析,进而构建舆情图谱,提出相关舆情引导建议。在以“东航MU5735坠机事件”为例的重大突发事件中,网民情感在不同舆情周期内主要伴随主流媒体的相关报道而产生变化,并因某些关键节点事件引发群众情感共鸣,产生截然不同的情绪。所构建的可视化舆情图谱可以为相应重大突发事件的舆情引导与监管提供借鉴。

By analyzing the distribution of netizens′emotions in different public opinion cycles and different themes in major emergencies,this paper constructs a public opinion map of major emergencies based on the dynamic analysis of emotions and themes,which can help guide public opinion in major emergencies.On the basis of introducing the construction method of the disaster event emotion dictionary,the LDA theme text mining method and the public opinion map construction method,data collection and processing are carried out based on Sina Weibo hot search topics.Next,based on the public opinion cycle division and topic classification,the dynamic collaborative analysis between netizen sentiment and public opinion cycle/topic is carried out,and then a public opinion map is constructed.Taking the“China Eastern Airlines MU5735 crash”as an example,the emotions of netizens mainly changed with the relevant reports of the mainstream media in different public opinion cycles,and because of some key node events,the emotional resonance of the masses was triggered,resulting in completely different emotions.In summary,the constructed visual public opinion map can provide reference for the public opinion guidance and supervision of corresponding major emergencies.