A COMPARATIVE STUDY OF FAKE NEWS DETECTION USING MACHINE LEARNING AND DEEP LEARNING TECHNIQUES

Abstract

The term "fake news" describes information that has been purposefully falsified or misrepresented and is presented as authentic journalism to mislead or control viewers. The rampant increase in misinformation and fake news across social platforms has created a multitude of problems across various spheres of society, affecting individuals, communities, and even global affairs. News plays an important role in our lives by providing us with current information from all over the world. With the rise in popularity of online news, there has been a complete change in how we consume news media. So, with the increasing popularity of social platforms along with the ease with which misinformation can be spread, a way to detect fake news has been paramount. In this study we have studied and analysed various papers related to detection of fake news using machine learning and deep learning models. The aim of the study is to evaluate and compare the performance of various algorithms in detecting fake news. The study compares six algorithms namely, Logistic Regression, XGBoost algorithm, Naive Bayes, Recurrent Neural Networks (RNNs), Long Short-term Networks (LSTMs), and Gated Recurrent Unit Networks (GRUs). In our research, we found that LSTM had the best performance for the WELFake dataset.