Multilingual Hate Speech Detection Using  Meta-Transfer-Learning

Author(s)	Kushal Sinha, Mohammed Basim Alam, Nandana Aravind, Priyanshu Hiranyareta, Surbhi Choudhary
Country	India
Abstract	—The fast expansion of online platforms created a record-setting rise in hate speech, undermining safe and wel coming online spaces. This project contributes a multilingual hate speech system based on meta-learning and transfer learning to achieve improved detection for languages. We employ a pre trained mBERT model for the English language and fine-tune it to the low-resource language of Hindi, Hinglish, Marathi and Bengali using meta-learning and transfer learning methods. With the incorporation of these practices, the system generalizes remarkably well to a new language given limited labeled exam ples, with impressive accuracy, scalability, and versatility. The proposed method greatly increases detection rates, reduces bias, and promotes healthy digital interactions ultimately leading to an improved safer online space. Besides, the framework of the system supports easy adaptation to other languages with small corpora.
Keywords	mBERT, Meta-Learning, Hate speech
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 16, Issue 2, April-June 2025
Published On	2025-05-06
DOI	https://doi.org/10.71097/IJSAT.v16.i2.4521
Short DOI	https://doi.org/g9hsqj

About IJSAT Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijsat.org

International Journal on Science and Technology