Fake News Identification Using Supervised Machine Learning Algorithms

Abstract

Fake news has emerged as a significant challenge in today's information-driven society, where misinformation can spread rapidly and have detrimental consequences. Detecting and combatting fake news is crucial in maintaining the integrity of news sources and ensuring the public's access to accurate and reliable information. Machine learning approaches have recently demonstrated the ability to recognize false news stories automatically based on their features and content. To identify fake news, this study compares and contrasts several machine learning (ML) methods, including Random Forest, Passive Aggressive Classifier (PAC), Multinomial Naive Bayes, SVC, Decision tree, Gradient boosting, XG Boost, and Logistic Regression. These algorithms are tested on WELFake_Dataset and the output received has shown a significant increase the accuracy and a decrease in the false rate.