The paper presents an approach to detecting violence-inciting texts on social media, a problem that has been exacerbated in low-resource settings due to sparse research and less data. The authors evaluated several BERT-based models and used an ensemble of these models for their final submission. The data for the study consisted of texts in the Bangla language, classified into one of three categories based on the types of violence-inciting texts. The authors’ submission ranked 10th in the final leaderboard of the shared task with a macro F1 score of 0.737.
Publication date: 1 Dec 2023
Project Page: Not mentioned
Paper: https://arxiv.org/pdf/2311.18778
