Agenda Setting on Social Media in Asia: A Text-Mining & Statistical Approach

• In Event: Political Communication Virtual Poster Session
  In Virtual Poster Session: Political Communication Virtual Poster Session I

Abstract

Politicians have been aware of the power of social media along with the rise of internet users in recent decades. Hence, politicians have tried to connect with supporters through different platforms. This article aims to generalize how political parties in East Asia draw supports on social media by agenda setting. Specifically, the study concerns whether the topic, the format (photo, video, and text), the timing (how far from the election and whether a political dispute just happened), and other factors would affect and how these factors affect the reactions of supporters on social media.

The paper attempts to crawl Facebook posts from social media account of political parties in three Asian societies, including Taiwan, Singapore, and Hong Kong, as English and Chinese text corpora are better developed, and they are all facing strong political tensions due to the US-China political disputes. The initial analysis on Taiwan’s case shows that the pro-democracy party’s posts can be divided into six catteries, including cross-strait businesses, party affairs, legislative affairs, national values, party promotion, and political system reform. On the other hand, the pro-Beijing party’s posts can be divided into cross-strait affairs, US-China relations, attacking others, among others. In addition to the difference of topics, other results also indicate that timings, formats, and topics do affect the reactions of supporters. This study also covers the cases of Singapore and Hong Kong.

In short, this article aims to figure out how political parties in East Asia draw supports on social media by agenda setting through statistical analysis and text mining.

Author

• Hugo Tai, University of Oxford