Addressing the critical issue of gender-based online abuse in the Bengali language, this repository presents a comprehensive study that leverages deep learning techniques, including BERT, mBERT, and BanglaBERT, for the identification of misogyny. We evaluate model performance using key metrics such as Accuracy, F1 Score, Precision, and Recall, shedding light on the effectiveness of these models in recognizing misogynistic language.
🌐 Our research hypothesis involves enhancing the mBERT model by incorporating linguistic and cultural diversity through multilingual training, including Bengali, Hindi, and English data. The goal is to improve the model's ability to detect misogyny in Bengali, potentially transcending language barriers and biases.
- Training and fine-tuning of BERT-based models for misogyny detection.
- Multilingual model enhancement for cross-lingual applicability.
- Model interpretability using LIME, providing insights into decision-making processes.
📈 The results offer valuable insights into the strengths and limitations of these models, contributing to the development of more effective tools for identifying online sexism. This study lays the foundation for future research on language-specific nuances and cross-lingual trends in the realm of gender-based abuse detection, ultimately working towards safer digital environments.
- Published in: 12th IEEE ComNetSat 2023
- DOI: 10.1109/COMNETSAT59769.2023.10420536