With the rise of social media,an increasing number of netizens express their opinions and emotions on these platforms,sometimes triggering negative group events with serious consequences.This paper takes the incident of a student's fall from Chengdu No.49 Middle School as an example,collecting relevant data from Weibo with the help of web crawlers.Long Short-Term Memory(LSTM)neural networks were employed for emotional classification,calculation of emotional heat,and division of the stages of emotional evolution.Analysis is also conducted on the emotions of netizens in public opinion stage by stage by using textual analysis including Latent Dirichlet Allocation.The research identifies five stages of emotional evolu-tion:the nascent phase,where netizens'emotions primarily focus on attention and information craving;the rising phase,characterized by complex emotions,unease and defiance;the peak phase,where emotions diver-sify and vigilance against rumors increases;the decline phase,which shifts towards rational reflection;the fading phase,where lessons from this incident are applied to the development of public opinion over subse-quent hot issues.
关键词
社交媒体/热点事件/情绪/情绪管理/情绪阶段/LSTM/LDA
Key words
social media/hot events/emotion/emotion management/emotional stages/LSTM/LDA
维普
万方数据