This study explores the use of BERT-based models for the detection of violence-inciting texts in social media, specifically in the Bangla language. The team evaluated several models and used an ensemble of these models for their final submission. The task was part of the First Workshop on Bangla Language Processing, where the team ranked 10th. The paper underscores the importance of efficient mechanisms to detect and curb the propagation of violence-inciting texts, especially in low-resource settings.
Publication date: 1 Dec 2023
Project Page: https://arxiv.org/abs/2308.11237
Paper: https://arxiv.org/pdf/2311.18778
