Brinjal: A Web-Plugin for Collaborative Hate Speech Detection

The proliferation of hate speech (HS) has compromised the safety and trustworthiness of the internet, exacerbating social divides by promoting hatred and discrimination. Although recent studies have produced guidelines and developed advanced technologies for the automated detection of HS, their efficacy and adaptability in real-world applications remain unclear. Furthermore, existing guidelines on what constitutes HS might not reflect the perspectives and beliefs of individuals and communities. This paper introduces Brinjal, a multifaceted web plugin designed for the collaborative detection of HS. Brinjal enables individuals to identify instances of HS and engage in discussions to verify such content, thereby enhancing the collective understanding of HS. Additionally, Brinjal serves as a practical platform for deploying and evaluating advanced HS detection models, facilitating user interaction and performance assessment. Lastly, Brinjal includes an analytical tool for analyzing HS, offering insights based on the crowdsourced instances and discussions about HS across various websites. The video demonstration of Brinjal can be viewed here: https://youtu.be/\_JxziIVWBO4. Disclaimer: This paper contains violent and discriminatory content that may be disturbing to some readers.