It has become clear that fake news is dangerous. Identifying fake news is a crucial step towards preserving the virtue and prosperity of society. Social media’s rising popularity has led to an increase in the spread of false information. There aren’t enough frameworks in place to deal with misleading news. There are many low-cost online news sources and it’s an easy access via social media. These are the reasons why there’s a spread of fake news. News Content is the only reason for the present fake news detection algorithms, also users’ previous posts or activities provide a lot of insights about their views on news and have a significant effect on false news identification. The proposed research seeks to investigate several machine learning approaches for the analysis and identification of false news. In order to identify the spread of fake news on social media, we compare various widely used machine learning methods, such as Naive Bayes and Multi Layer Perceptron Classifiers, in this study. In this work, using solely text data, we develop a number of machine learning methods using the WELFake dataset.

Index Terms—fake news detection, machine learning, text analysis, naive bayes, multi layer perceptron