Evaluation of Machine Learning Models for Afan Oromo Fake News Detection

The rapid proliferation of fake news, particularly in low-resource languages like Afan Oromo, poses significant challenges to information integrity and societal trust. This study evaluates the performance of various machine learning models for detecting fake news in Afan Oromo, a language with limited digital resources and linguistic tools. Several supervised learning algorithms, including Support Vector Machines (SVM), Random Forest, Logistic Regression, and deep learning-based approaches such as Long Short-Term Memory (LSTM) networks, are trained and tested on a curated dataset of Afan Oromo news articles. The dataset is preprocessed using tokenization, stemming, and feature extraction techniques tailored to the linguistic characteristics of Afan Oromo. Model performance is assessed using metrics such as accuracy, precision, recall, and F1-score. Results indicate that deep learning models, particularly LSTM, outperform traditional machine learning algorithms in capturing the contextual nuances of Afan Oromo text. However, the study also highlights the challenges posed by the language's morphological complexity and the scarcity of annotated data. This research contributes to the growing body of work on fake news detection in low-resource languages and provides insights into the development of robust models for Afan Oromo and similar languages.