微博涉核舆情的情感分析与公众传播研究

The Sentimental Analysis and Public Communication of Microblog’s Nuclear Related Public Opinion

[目的/意义]涉核话题近年来随着核产业的快速发展备受各界关注,负面的涉核网络舆情的公众传播是当前政府面临的新挑战,而微博作为网络舆情传播的重要平台,通过对微博涉核内容的情感倾向分析有助于深入探究负面舆情的发展传播规律。[方法/过程]以微博涉核话题为数据源,提出了多任务多输入的协同情感分类模型,对微博内容从时间、地域、用户关系三方面进行了情感倾向分析。[结果/结论]通过使用多任务多输入的协同情感分类模型,对微博这种长度较短且网络用语较多的文本进行分析,并综合数据分析得出了公众涉核舆情的发展传播形势。

[Purpose/significance]In recent years,with the rapid development of nuclear industry,nuclear related topics have attracted attention from all aspects of life.The public communication of negative nuclear related network public opinion is a new challenge facing the current government.As an important platform of Internet public opinion communication,Microblog,is selected to conduct analysis of the emotional tendency as far as nuclear related content is concerned.It helps insightful exploration of the development and communication law of negative public opinion.[Method/process]Based on Microblog’s nuclear related topics,this paper proposes a multi-task and multi-input collaborative emotion classification model,and analyzes the emotion tendency of Microblog’s content from three aspects which are time,region and user relationship.[Result/conclusion]By using the multi-task and multi-input collaborative emotion classification model,the text of micro-blog,which is shorter in length and has more network language,is analyzed,and the comprehensive data analysis helps to understand the development and dissemination of nuclear related public opinion.