面向公安微博的用户评论情感分析及反馈研究

Research on Sentiment Analysis and Feedback of User′s Comments on Public Security Microblog

为分析公安微博在热点舆情事件的处置和应对过程,通过网络爬虫获取公众反映激烈的“女子半夜遭殴打”事件中的公安微博评论,利用Word2vec与SVM方法相结合的方式,对评论文本数据进行情感极性分析,并对公众情绪在应对过程中的变化特征进行研究,发现使用公安微博及时传递信息、回应公众关切能有效控制负面情绪的爆发。故结合自动控制原理,构建公安微博舆情反馈模型,能促进微博警务工作的改进,为公安微博处理公众热点事件提供借鉴与参考。

In order to analyze the handling and response process of public security microblogs in hot public opinion events,public security microblog comments in the event of"a woman was beaten in the middle of the night"which received a heated reaction from the public through web crawlers,word2vec and SVM methods were used to conduct a sentiment analysis of these comments,and the changing characteristics of public emotions in the process of coping was studied.It was found that the use of public security microblogs to timely transmit information and respond to public concerns can effectively control the outbreak of negative emotions.Therefore,combined with the principle of automatic control,a public security microblog public opinion feedback model was constructed to promote the improvement of microblog police work and provide reference for public security microblogs to handle public hot events.