Fake News Detection Using Deep Learning

Abstract

The unexpected surge in social media has given rise to distribution of unlimited content over the internet. The social media platforms behave like a broadcasting medium these days. Increase in social media users over the past decade has given these platforms a liberty to own and entertain their audience. Users can access daily news through these platforms. Various channels (news related) have their official accounts which usually post news online. Publicizing and forwarding the content online is very easy and free. This is one of the reasons why fake news is circulated on the web. Platforms like Facebook, Instagram, and twitter are used by millions of people every day and therefore, data is generated with a great velocity. Huge amount of data out there exists that is difficult to process and classify into false data. Some users intentionally or unintentionally communicate this kind of stuff online. Fake news or fake information can be of any form i.e, post, image, video, audio, etc. Proliferation of this kind of data amongst the people may mislead them to believe something that is not true. The widespread usage of social media plays a vital role in setting up the mindsets of the people and their actions in the real world. The free will provided by these platforms(social media), allows people to upload it easily and hence, it is circulated among a huge number of masses. The credibility of any content being uploaded on social media platforms is still a quest and the major reason for fake news spread. This is the very motivation of this study. This research study is a systematic review of the recent work performed in detecting the fake news using deep learning. Herein, we perform a task for improving the performance of deep learning models and also we implement the graph neural network for the identification of the false news.