基于情感分析的网络民愤研究——以长春长生问题疫苗事件为例

Sentiment analysis of public indignation in social media:a problematic vaccine incident as an example

以长春长生问题疫苗事件中的网民评论为研究样本,采用双向长短期记忆模型识别出网络民愤,再基于TFIDF算法对网络民愤进行主题词、主题句的提取,并运用话语分析进行主题归纳。研究发现,网络民愤的主题围绕疫苗的非法生产和相关部门的责任缺失展开,这反映出公众对自身健康的担忧,对企业道德失范的焦虑,以及对党和政府肃清问题疫苗乱象的期待。利用这种情感分析方法能够真实地反映出民愤产生的原因及网民现实的社会心理状态,对于实现更精准的舆情分析和减少社会冲突具有重要意义。

This paper,taking the comments of netizens in Changchun Changsheng Problematic Vaccine Event as research sample,uses two-way long-term and short-term memory models to identify netizens’indignation,and then summarizes the themes of netizens’indignation based on TFIDF algorithm.The study found that the cyber-public anger centered on their dissatisfaction with the illegal production and sale of vaccines,and the lack of responsibility of relevant departments,which reflects the public's worries about their life and health,the panic of the moral decay of the company,and the social psychology about the handling of the vaccination disorder by the CPC Party and the government.This application of sentiment analysis can truly reflect the causes of the public indignation and its realistic social and psychological situations,which is of great significance for achieving more accurate public opinion analysis and reducing social conflicts.