2019-01-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/652379摘要：目前關於中國網路言論審查的文獻似乎均同意一個主張，即中國政府為了提升治理品質，刻意地允許某些批評政府的訊息，但卻刪除那些可能會煽動集體行動的言論。本計畫挑戰此一觀點。我們提出了一個咎責迴避的理論，即在某些情況下，威權統治者可以容忍某些具有集體行動內涵的訊息。本計畫採用文字探勘技術，分析在重大公共衛生事件發生時（包括2016年魏澤西事件與2018年疫苗醜聞），新浪微博內的貼文內容。此一研究有助於說明中國政府如何引導風向，將公眾批評轉向民間人士或企業。<br> Abstract: The current literature on China`s internet censorship seems to agree on one proposition: In order to improve its governance, the Chinese government strategically allows the messages criticizing it, while censors those having the potential to incite collective actions. This project takes issue with this. We propose a theory of scapegoating that, under certain conditions, it pays for the dictator to allow some messages with collective action potential to be posted. Empirically, we employ text mining technics to analyze the posts on Sina Weibo during several public health incidents including the 2016 Wei Zexi death and the 2018 vaccine scandal. These studies help to show how the Chinese government deflects public criticism to non-government actors.文字探勘言論審查新浪微博咎責迴避公共衛生Text-miningCensorshipSina WeiboDeflectionPublic Health