Racism detection using deep learning techniques

Abstract

With the pervasive role of social media in the socio-political landscape, various forms of racism have arisen on these platforms. Racism can manifest in various forms on social media, both concealed and overt. It can be hidden through the use of memes or exposed through racist comments made using fake profiles to spread social unrest, violence, and hatred. Twitter and other social media sites have become new settings in which racism and related stress appear to be thriving. Racism also spread based on characteristics including dialect, faith, and tradition. It has been determined that racial animosity on social media poses a serious threat to political, socioeconomic, and cultural equilibrium and has even put international peace at risk. Therefore, it is crucial to monitor social media as the primary source of racist opinions dissemination and to detect and block racist remarks in a timely manner. In this study, we aim to detect tweets containing racist text by performing sentiment analysis using both ML and DL algorithms. We will also build a webpage using Flask framework and SQLite for users to interact with the model.