Political mud slandering and power dynamics during Indian assembly elections

Political parties often engage in verbal swordplay, which worsens during elections. Free and fair elections are the pillar of a democratic society. Come election season, all media outlets, including the Internet, are buzzing with politically-charged content. We curated 46 k posts from Twitter between January and March, 2022 to examine political content during the Indian assembly elections of February 2022. We begin our analysis by manually labeling 1.7 k posts for different forms of attack and employ the annotation to examine political attacks against defamatory hashtags and name-calling. It anecdotally lends itself to the target’s online reputation. Similarly, we quantify the power dynamics of self-promotion and negation pinned on the ruling party before and after elections. To aid large-scale analysis, we obtain pseudo-labels for the rest of the dataset via training a political attack detector for the Indic setting. Subsequently, we observe that the patterns detected via manual annotations hold at scale too. Our analyses and findings aim to educate the citizenry about the quality of political discourse on Indian Twitter.