RETRACTED ARTICLE: Systematic approach for fake news detection using machine learning

Abstract

Thanks to the advent of W3C and the rapid proliferation of social media platforms (Facebook, Instagram, Twitter, etc.), an unprecedented level of knowledge sharing has become possible. Users are producing more n more information and it is being circulating to a large number of people. The produced information may not be correct all the time. Even subject matter experts need to judge an array of variables before evaluating to check the trustworthiness. In this study, we propose to use machine learning (ML) to automate message classification. Our research examines various criteria that can be used to differentiate between genuine materials and counterfeits. These qualities are used for training machine learning algorithms, and we then use datasets from the real world to evaluate how well they work. The result shows that Xgboost model gives the highest accuracy for fake news detection in both Tf-Idf and BOW feature extraction techniques.SVM and Multinomial Naïve Base models are the most underperforming models in Tf-Idf and BOW respectively.