NSU_MILab at BLP-2025 Task 1: Decoding Bangla Hate Speech: Fine-Grained Type and Target Detection via Transformer Ensembles

Md. Mohibur Rahman Nabil, Muhammad Rafsan Kabir, Rakib Islam, Fuad Rahman, Nabeel Mohammed, Shafin Rahman


Abstract
This paper describes our participation in Task 1A and Task 1B of the Task 1A and Task 1B of the BLP Workshop, focused on Bangla Multi-task Hatespeech Identification. Our approach involves systematic evaluation of four transformer models: BanglaBERT, XLM-RoBERTa, IndicBERT, and Bengali Abusive MuRIL. To enhance performance, we implemented an ensemble strategy that averages output probabilities from these transformer models, which consistently outperformed individual models across both tasks. The baseline classical methods demonstrated limitations in capturing complex linguistic cues, underscoring the superiority of transformer-based approaches for low-resource hate speech detection. Our solution initially achieved F1 scores of 0.7235 (ranked 12th) for Task 1A and 0.6981 (ranked 17th) for Task 1B among participating teams. Through post-competition refinements, we improved our Task 1B performance to 0.7331, demonstrating the effectiveness of ensemble methods in Bangla hate speech detection.
Anthology ID:
2025.banglalp-1.40
Volume:
Proceedings of the Second Workshop on Bangla Language Processing (BLP-2025)
Month:
December
Year:
2025
Address:
Mumbai, India
Editors:
Firoj Alam, Sudipta Kar, Shammur Absar Chowdhury, Naeemul Hassan, Enamul Hoque Prince, Mohiuddin Tasnim, Md Rashad Al Hasan Rony, Md Tahmid Rahman Rahman
Venues:
BanglaLP | WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
461–467
Language:
URL:
https://aclanthology.org/2025.banglalp-1.40/
DOI:
Bibkey:
Cite (ACL):
Md. Mohibur Rahman Nabil, Muhammad Rafsan Kabir, Rakib Islam, Fuad Rahman, Nabeel Mohammed, and Shafin Rahman. 2025. NSU_MILab at BLP-2025 Task 1: Decoding Bangla Hate Speech: Fine-Grained Type and Target Detection via Transformer Ensembles. In Proceedings of the Second Workshop on Bangla Language Processing (BLP-2025), pages 461–467, Mumbai, India. Association for Computational Linguistics.
Cite (Informal):
NSU_MILab at BLP-2025 Task 1: Decoding Bangla Hate Speech: Fine-Grained Type and Target Detection via Transformer Ensembles (Nabil et al., BanglaLP 2025)
Copy Citation:
PDF:
https://aclanthology.org/2025.banglalp-1.40.pdf