ADVANCED MACHINE LEARNING FRAMEWORK FOR IDENTIFYING AND MITIGATING FAKE NEWS AND MISINFORMATION PROPAGATION ON SOCIAL MEDIA PLATFORMS

Abstract

Fake news has become such a serious problem to everyone of all ages and backgrounds as it can deceive. The demand for accurate and reliable methods to detect misinformation has increased due to the rising reliance of individuals on digital platforms for entertainment, news, products, and services. Recent research efforts have tackled this problem, by focusing on developing actionable tactics to prevent the spread of false- information. This paper therefore proposes an ensemble machine learning method by combining Multinomial Naive Bayes, Random Forest, and Logistic Regression to enhance the accuracy and performance of false news detection. The ensemble model to take advantage of each special method and minimize its shortcomings. At a very high false positive rate accuracy of 96% and an F1 score of 95%, experimental data show that our suggested approach performs noticeably better than individual classifiers and other models already in use.