Supervised Machine Learning Algorithms for Fake News Detection

Abstract

In our modern era where the Internet is ubiquitous, everyone consumes various informations from the online resources. Along with the use of a huge amount of social media, news spread rapidly among the millions of users within a short interval of time. However, the quality of news on social media is lower than the traditional news outlets; the main reason behind that is the large amount of fake news. So in this paper, we have explored the application of machine learning techniques to identify the fake news. We have developed two models with the help of support vector machine, random forest, logistic regression, naive Bayes, and k-nearest neighbor machine learning algorithms, and this method is compared in terms of accuracy. A model focuses on identifying the fake news, based on multiple news articles (headline) and Facebook post data which gather informations about user social engagement. We achieved maximum classification accuracy of 98.25% (logistic regression) for a dataset A and 81.40% (KNN) accuracy for a dataset B.