【摘要】 近年来,反腐成为中国政治治理的重要议题。既有研究发现公众在该类议题的讨论中,情绪化表达时常代替理性的政治参与。然而,公众情绪如何影响,有哪些类型及产生原因等方面的细化研究较少。以情绪归因理论为分析工具,运用自然语言处理技术对2013-2016年中影响力较大的32个反腐案件中的网络情绪进行细粒度的情绪及其原因识别,可发现:在反腐议题的网络讨论中,情绪化表达,特别是低唤醒度的负面情绪表达最为显著;网民的情绪化表达更倾向于内部归因,而外部归因的正面情绪显著高于内部归因;宏观的社会环境和微观的事件特征对反腐议题中的网络情绪归因存在显著影响。更多还原

【Abstract】 In recent years,anti-corruption has become an important issue in China’s political governance. Based on the emotional assessment model and attribution of emotion theory,this paper adopts natural language processing technology to carry out a fine-grained sentiment and cause identification of a large-scale online expression text in 32 anti-corruption cases with high impact in 2013-2016. The study found that in the online discussion on anti-corruption issues,negative emotion with low arousal is most prominent; Netizens’ emotional expression is more inclined to internal attribution,and positive emotion of external attribution is significantly higher than internal attribution; Moreover,social environment and characteristics of cases have significant impact on online expression and attribution. However,after controlling these factors,the valence of online emotion is almost stable in different attribution. This study further discusses and explains the underlying causes and possibilities of these emotional expression features from the ‘observer ’attribution tendency,‘result dependence’emotion,and ‘attribution dependence’emotion.更多还原