Advancing Social Stability: A Comparative Assessment of Fake News Detection through Machine Learning and Deep Learning Methodologies

Abstract

Ensuring the credibility of information in the digital era requires effective detection of fake news. This study compares multiple machine learning and deep learning models, including logistic regression, random forest, decision tree, gradient boosting, CNN, RNN, and LSTM. Performance was evaluated using accuracy, precision, recall, and F1-score. Results show that logistic regression achieved the highest performance (92.19% accuracy, F1 = 0.92), outperforming even advanced deep learning methods. These findings suggest that simple, efficient models can be highly reliable for fake news detection in real-world applications.