We have witnessed the social media boom in China last decade. There are 0.8 billion Chinese internet users (which outnumbered the population of Europe) today, and digital platforms such as Weibo and WeChat has changed the way people interact with others. But the activities in those platforms are neither free from monitored nor without manipulation by the government. Since the free flow of information would cause destabilization to the Chinese Communist Party's autocratic regime, CCP has built the Great Firewall of China (GFW) to regulate domestic internet. According to Gary King's works, the GFW not only blocks foreign websites Google or Facebook but also helps the government to censor certain "sensitive" informations, especially those called for collective action. Another hidden tool for the Chinese government to manipulate trends are thousands of "50 cent Army", who post tons of information to promoted pro-government narratives. This paper attempts to discover the interaction of online civil society and autocratic regulatory apparatus, by focusing on one special issue: Taiwan.
"Taiwan issue", which is two-way politically sensitive on popular social media networks. On one side, it is apodictic truth to claim on Taiwan as a part of China on the internet. Moreover, when there is any trace of "pro-independent" narrative, it often sparks volcanic nationalism trend over social media. But on the other side, as a democratic regime, Taiwan would regularly hold free and fair elections. To China's cruel regulator, it would be reasonable to put down online discussions over Taiwan's election, since that information might easily let people wonder whether they could have their own. Then there is the puzzle: should government propagandize over Taiwan's election as unacceptable "pro-independent" event and call for online-bullying, or should the government put that information down as they could remind people how "Chinese democracy" could be one day?
To solve the puzzle, we first scrapped down all web pages on the Chinese digital platform which contained "Taiwan" during the 2014 Taiwan general election period. Second, using machine learning text-analysis and network analysis, we categorized those web pages as political or non-political, calculate the frequencies on a daily basis. We also labeled emotions within each thread or news, and construct opinion-network model to find whether the trends are manipulated by government or evolve from civil society. The results showed that online discussion over Taiwan was sky-high during the election period. There is little censorship over those discussions since most of the opinion are pro-government. Also, there exist some spin control exercises trying to neutralize the pro-democratic discussion. The state media even shut the news coverages over Taiwan's election one day after it. Our findings reveal the Chinese government's two-pronged approach to regulating a sensitive issue online.
But there is a silver lining on netizen's side. Despite the government tried to cool the discussion, there is still hot debate over Taiwan's democratic election. In addition, even most of the internet users dislike Taiwan's pro-independent election outcome, there are still lots of opinions appraise democratic election as favorable.
