ReBERT at HSD-2Lang 2024: Fine-Tuning BERT with AdamW for Hate Speech Detection in Arabic and Turkish

Yagci U. U., Kolcak A. E., Iscan E.

7th Workshop on Challenges and Applications of Automated Extraction of Socio-Political Events from Text, CASE 2024, St. Julian's, Malta, 22 March 2024, pp.195-198, (Full Text) identifier

  • Publication Type: Conference Paper / Full Text
  • City: St. Julian's
  • Country: Malta
  • Page Numbers: pp.195-198
  • Middle East Technical University Affiliated: Yes

Abstract

This research tackles the issue of detecting hate speech in Arabic and Turkish languages by utilizing pre-trained BERT models, namely TurkishBERTweet and Arabertv02-twitter. These models are enhanced through a comprehensive hyperparameter search to improve their performance. Our classifiers excelled in the HSD-2Lang 2024 contest, with the Turkish model placing second in Subtask A and the Arabic model first in Subtask B on the private leaderboard. Both models also ranked first on the public dataset. These results demonstrate the efficacy and adaptability of our approach in addressing the evolving challenges of hate speech detection in multilingual contexts.