Using Social Media Categorical Reactions as a Gateway to Identify Hate Speech in COVID-19 News

Abstract

The spread of COVID-19 news on social media provided a particularly prolific ground for emotional commotion, disinformation and hate speech, as uncertainty and fear grew by the day. In this paper, we examine the media coverage of the COVID-19 outburst in Portugal (March–May 2020), the subsequent emotional engagement of audiences and the entropy-based emotional controversy generated as a gateway to detect the presence of hate speech, using computer-assisted qualitative data analysis (CAQDAS) embedded in a cross-sectional descriptive methodology. Our results reveal that negative and volatile categorical emotions (“Angry”, “Haha” and “Wow”) serve as main engines for controversy, and that controversial news have the highest sharing ratio. Moreover, using a small sample of the most controversial news with the highest overall emotional engagement, we establish a relation between the entropy-based emotional controversy obtained from Facebook’s click-based reactions and the presence of cultural and ethnic hate speech, plausibly confirming the click-based categorical emotions as a gateway to hatred comment pools. In doing so, we also reveal that negative emotions alone do not always indicate the presence of hate speech, which may sprout in seemingly humorous social media posts where irony proliferates, and negativity is not apparent. This work adds to the literature on social media categorical emotion detection and its implications for the detection of hate speech.