HausaNLP at SemEval-2025 Task 11: Hausa Text Emotion Detection

TL;DR

This paper describes a transformer-based approach using AfriBERTa for multi-label emotion detection in Hausa, achieving over 73% F1-score, demonstrating effectiveness in low-resource language NLP tasks.

Contribution

It introduces fine-tuning of AfriBERTa for Hausa emotion detection, a low-resource language, showing promising results for NLP in underrepresented languages.

Findings

Validation accuracy of 74.00%

F1-score of 73.50%

Effective use of transformer models for Hausa NLP

Abstract

This paper presents our approach to multi-label emotion detection in Hausa, a low-resource African language, for SemEval Track A. We fine-tuned AfriBERTa, a transformer-based model pre-trained on African languages, to classify Hausa text into six emotions: anger, disgust, fear, joy, sadness, and surprise. Our methodology involved data preprocessing, tokenization, and model fine-tuning using the Hugging Face Trainer API. The system achieved a validation accuracy of 74.00%, with an F1-score of 73.50%, demonstrating the effectiveness of transformer-based models for emotion detection in low-resource languages.