Hybrid Machine Learning Models for Enhanced Fake News Detection

Abstract

This paper states a comprehensive approach to developing a robust fake news recognition system by leveraging advanced machine learning techniques. The increasing proliferation of fake news on digital platforms has made it essential to create systems that can automatically and accurately distinguish between factual and fabricated news articles. Our proposed system focuses on classifying news articles as fake or real based on their textual content, employing a category of machine learning classifiers such as Logistic Regression, Decision Trees, Gradient Boosting, and Random Forests. To extract meaningful features from the text, we utilized Term Frequency-Inverse Document Frequency vectorization, which converts textual data into numerical vectors suitable for analysis. The classifiers were trained and rigorously evaluated on a labeled dataset of fake and real news articles, and the results demonstrated significant accuracy across all models. Moreover, the system includes a manual testing function that enables real-time input and classification of news articles, showcasing its practical application in identifying fake news with high reliability. This paper highlights the effectiveness of traditional machine learning models and demonstrates the importance of combining them with feature engineering techniques to combat misinformation in the digital age.