Abstract:
India is a diverse country with very active politics. In recent times (2014-2021), a lot of polit- ical activities have stirred activities in the online space. These political activities range from Elections to Protests against policies. So, we explore the online space, Twitter, in speci c un- der two settings: the way they are used by politicians in general and the way they are used by public during protests. The rst study of studying the way politicians use social media is done via Framing. Framing is the way of portraying thoughts and opinions in order to sway the public opinion in one di- rection. The study is conducted in two major languages used in Indian political online space: Hindi and English. We tried to study how these frames vary across political parties and is- sues. The second study of user behaviour during protests is done using unsupervised clustering of users to nd the diferent groups they tend to form during the protests. This study is done at user level using the unsupervised stance detection method. CAA-protest and Farmer-protest are the speci c datasets used for this work.
