Theme clustering map of network public opinion and sentiment evolution in public security emergencies based on LDA model
In view of the hot issues of large-scale discussion among netizens after public security emergencies,this paper car-ries out network public opinion analysis based on LDA model.Taking an aircraft crash as an example,this paper constructs a text topic clustering map,and uses point mutual information to extract high-frequency words for semantic network analysis,and then analyzes the relationship between event elements through the semantic association of keywords.Finally,the emotional evolution characteristics of netizens are analyzed according to emotional statistics and heat evolution.The research shows that the topic clus-tering map and sentiment analysis are helpful to sort out the event context and timely prevent the invisible risk of network public opinion in the event,and provide reference for exploring the universal countermeasures of public opinion governance.
万方数据
