@inproceedings{hasan-etal-2026-llm,
title = "{LLM}-Based Multi-Task {B}angla Hate Speech Detection: Type, Severity, and Target",
author = "Hasan, Md Arid and
Alam, Firoj and
Hossain, Md Fahad and
Naseem, Usman and
Ahmed, Syed Ishtiaque",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.1565/",
pages = "33962--33980",
ISBN = "979-8-89176-390-6",
abstract = "Online social media platforms have become central to communication and information exchange, however, they also serve as fertile ground for hate speech, offensive language, and bullying targeting individuals and communities. Such content undermines online safety and inclusion, underscoring the need for reliable detection systems{---}especially in low-resource languages with limited moderation tools. For Bangla, existing work provides valuable resources and models, however, they are mostly single-task (e.g., binary hate/offense) with narrow coverage of key dimensions such as type, severity, and target. We address these gaps by introducing *the first multi-task* Bangla hate-speech dataset, *BanglaMultiHate*, one of the largest manually annotated dataset to date. Using this resource, we performed a comparative study across different baselines, monolingual pretrained models, and LLMs under zero-shot, few-shot, and LoRA fine-tuning settings. Our findings show that while LoRA-tuned LLMs rival BanglaBERT, culturally grounded pretraining remains crucial for robust performance. Overall, *BanglaMultiHate* establishes a stronger benchmark for hate speech detection in low-resource contexts. All data and scripts are released for reproducibility."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hasan-etal-2026-llm">
<titleInfo>
<title>LLM-Based Multi-Task Bangla Hate Speech Detection: Type, Severity, and Target</title>
</titleInfo>
<name type="personal">
<namePart type="given">Md</namePart>
<namePart type="given">Arid</namePart>
<namePart type="family">Hasan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Firoj</namePart>
<namePart type="family">Alam</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Md</namePart>
<namePart type="given">Fahad</namePart>
<namePart type="family">Hossain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Usman</namePart>
<namePart type="family">Naseem</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Syed</namePart>
<namePart type="given">Ishtiaque</namePart>
<namePart type="family">Ahmed</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Online social media platforms have become central to communication and information exchange, however, they also serve as fertile ground for hate speech, offensive language, and bullying targeting individuals and communities. Such content undermines online safety and inclusion, underscoring the need for reliable detection systems—especially in low-resource languages with limited moderation tools. For Bangla, existing work provides valuable resources and models, however, they are mostly single-task (e.g., binary hate/offense) with narrow coverage of key dimensions such as type, severity, and target. We address these gaps by introducing *the first multi-task* Bangla hate-speech dataset, *BanglaMultiHate*, one of the largest manually annotated dataset to date. Using this resource, we performed a comparative study across different baselines, monolingual pretrained models, and LLMs under zero-shot, few-shot, and LoRA fine-tuning settings. Our findings show that while LoRA-tuned LLMs rival BanglaBERT, culturally grounded pretraining remains crucial for robust performance. Overall, *BanglaMultiHate* establishes a stronger benchmark for hate speech detection in low-resource contexts. All data and scripts are released for reproducibility.</abstract>
<identifier type="citekey">hasan-etal-2026-llm</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.1565/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>33962</start>
<end>33980</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T LLM-Based Multi-Task Bangla Hate Speech Detection: Type, Severity, and Target
%A Hasan, Md Arid
%A Alam, Firoj
%A Hossain, Md Fahad
%A Naseem, Usman
%A Ahmed, Syed Ishtiaque
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F hasan-etal-2026-llm
%X Online social media platforms have become central to communication and information exchange, however, they also serve as fertile ground for hate speech, offensive language, and bullying targeting individuals and communities. Such content undermines online safety and inclusion, underscoring the need for reliable detection systems—especially in low-resource languages with limited moderation tools. For Bangla, existing work provides valuable resources and models, however, they are mostly single-task (e.g., binary hate/offense) with narrow coverage of key dimensions such as type, severity, and target. We address these gaps by introducing *the first multi-task* Bangla hate-speech dataset, *BanglaMultiHate*, one of the largest manually annotated dataset to date. Using this resource, we performed a comparative study across different baselines, monolingual pretrained models, and LLMs under zero-shot, few-shot, and LoRA fine-tuning settings. Our findings show that while LoRA-tuned LLMs rival BanglaBERT, culturally grounded pretraining remains crucial for robust performance. Overall, *BanglaMultiHate* establishes a stronger benchmark for hate speech detection in low-resource contexts. All data and scripts are released for reproducibility.
%U https://aclanthology.org/2026.acl-long.1565/
%P 33962-33980
Markdown (Informal)
[LLM-Based Multi-Task Bangla Hate Speech Detection: Type, Severity, and Target](https://aclanthology.org/2026.acl-long.1565/) (Hasan et al., ACL 2026)
ACL
- Md Arid Hasan, Firoj Alam, Md Fahad Hossain, Usman Naseem, and Syed Ishtiaque Ahmed. 2026. LLM-Based Multi-Task Bangla Hate Speech Detection: Type, Severity, and Target. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 33962–33980, San Diego, California, United States. Association for Computational Linguistics.