Fake News Detection using Machine Learning: Integrating FakeBERT Classification, Style Analysis, and Credibility Verification

Fake news refers to false information spread to incite hate, propaganda, or seek attention. It can quickly spread through social media, significantly impacting society. This research aims to improve information accuracy by developing robust Machine Learning models for detecting fake news to enhance awareness of misinformation through integration into a web-based application and provides insights for further research in fake news detection. This is achieved by integrating three approaches: FakeBERT Classification, Style Analysis, and Credibility Verification. The study utilizes the ISOT Fake News Dataset from the University of Victoria, verified by Politifact.com, which primarily consists of political news articles. The study compares three machine learning models: Decision Tree, Random Forest, and XGBoost. The results indicate that Random Forest achieves better precision and accuracy. Additionally, the model is fine-tuned, achieving 89% precision for the Fake class and 86% for the True class on the training set, and 96% for the Fake class and 86% for the True class on the testing set.