Profiling Social Media Campaigns and Political Influence: The Case of Pakistani Politics

Social networking websites have been widely used for political awareness and campaigns. With the increasing reliance of political parties and supporters on social media, interesting patterns can be extracted by analyzing social media profiles. The previous approaches perform political analysis used social media data mainly focused on analyzing only textual data for the two-party system. However, the social media engagement, i.e. posts reaction, of political supporters can be used to identify the most influential political party. Therefore, to explore the content of social media political campaigns and to identify the most influential political party using social engagement in a multi-party system, we extracted Facebook data using Facebook Graph API for the three most popular political parties in Pakistan, i.e. PTI, PML-N and, PPP. The results reported that PTI relies more on the English language when compared with PML-N and PTI. However, video content is most prominently used by all three parties. The ratio of reactions on Facebook posts varies, PTI has received the highest number of "LIKE" and "SHARE", PML-N, and PPP has received the major number of "HAHA" and "SAD". PTI appears as the most influential party in comparison with PML-N and PPP. The influence score for each party appeared in accordance with the results of Pakistan's General Elections 2018. Hence, the proposed influence score irrespective of party-system can be used by researchers to improve the accuracy of electoral prediction. The insights highlighted by this study can also be used by political parties to improve their social media campaigns for increasing social influence.