Dataset Creation and Baseline Models for Sexism Detection in Hausa

2025-10-30

AI Summary

To address the scarcity of gender bias detection research for low-resource languages like Hausa, this study constructs the first high-quality, culturally grounded Hausa annotated dataset for gender discrimination detection, leveraging community collaboration, qualitative coding, and data augmentation. Methodologically, it introduces a two-stage user study—including in-depth interviews with native speakers and iterative annotation feedback—to systematically identify and model culture-specific linguistic phenomena (e.g., clarification requests, idiomatic misinterpretations). It further benchmarks traditional machine learning and multilingual pretrained models under few-shot settings. Results show that while multilingual models exhibit baseline discrimination detection capability, their lack of cultural contextualization leads to substantial false positives. This work fills a critical gap in computational gender bias research for Hausa and proposes a “culture-embedded data construction” paradigm—a reusable methodological framework for bias detection in low-resource languages.

Technology Category

• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Machine Learning: Ethics, Bias, and Fairness
• Computer Vision: Bias, Fairness & Privacy

Application Category

• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Sexism reinforces gender inequality and social exclusion by perpetuating stereotypes, bias, and discriminatory norms. Noting how online platforms enable various forms of sexism to thrive, there is a growing need for effective sexism detection and mitigation strategies. While computational approaches to sexism detection are widespread in high-resource languages, progress remains limited in low-resource languages where limited linguistic resources and cultural differences affect how sexism is expressed and perceived. This study introduces the first Hausa sexism detection dataset, developed through community engagement, qualitative coding, and data augmentation. For cultural nuances and linguistic representation, we conducted a two-stage user study (n=66) involving native speakers to explore how sexism is defined and articulated in everyday discourse. We further experiment with both traditional machine learning classifiers and pre-trained multilingual language models and evaluating the effectiveness few-shot learning in detecting sexism in Hausa. Our findings highlight challenges in capturing cultural nuance, particularly with clarification-seeking and idiomatic expressions, and reveal a tendency for many false positives in such cases.

Problem

• Creating the first annotated dataset for detecting sexism in Hausa language
• Addressing cultural nuances in sexism expression through community engagement
• Evaluating machine learning models for low-resource language sexism detection

Innovation

• Developed first Hausa sexism dataset through community engagement
• Conducted two-stage native speaker study for cultural nuances
• Evaluated few-shot learning with multilingual models for detection