Cyberbullying Detection in Hausa Language on X Social Medium Using Machine Learning.

The rise of social media has intensified cyberbullying, impacting users across diverse languages, yet low-resource languages like Hausa lack effective detection tools. With over 100 million Hausa speakers predominantly in Nigeria and Niger, addressing this gap is crucial for fostering safer online environments. This study aims to develop a machine learning model to detect cyberbullying in Hausa tweets on X (formerly Twitter). The methodology involved collecting and annotating Hausa-language tweets, preprocessing the data through cleaning and removal of stopwords using Natural Language Tokenization (NLTK) Library, and extracting features using TF-IDF technique. Multiple classifiers, including Support Vector Machine (SVM), XGBoost, and Logistic Regression, were trained and evaluated based on accuracy, precision, and recall. Results showed that the SVM outperformed others with an accuracy of 0.98, followed by Random Forest 0.9794, then XGBoost 0.9769, while Logistic Regression had the lowest accuracy 0.9527. The findings demonstrate that culturally-sensitive, language-specific models can enhance cyberbullying detection in Hausa, contributing to safer digital spaces and providing a basis for further research in low-resource language NLP applications.