Identification of Fake News Using Machine Learning

Fake news has been a problem ever since the internet boomed. The very network that allows us to know what is happening globally is the perfect breeding ground for malicious and fake news. Combating this fake news is important because the world's view is shaped by information. People not only make important decisions based on information but also form their own opinions. If this information is false it can have devastating consequences. Verifying each news one by one by a human being is completely unfeasible. This paper attempts to expedite the process of identification of fake news by proposing a system that can reliably classify fake news. Machine Learning algorithms such as Naive Bayes, Passive Aggressive Classifier and Deep Neural Networks have being used on eight different datasets acquired from various sources. The paper also includes the analysis and results of each model. The arduous task of detection of fake news can be made trivial with the usage of the right models with the right tools.