Defining News Authenticity on Social Media Using Machine Learning Approach

Abstract

Social network and online news media are becoming popular in today’s era. Due to low cost, easy access and rapid diffusion, social media platform becomes a source to distribute false information. Fake news propagation on social media can cause serious negative effects on human society especially in politic, reputation and finance. So, automatic fake news detection plays a vital role to robust news media platform on social network. Defining news authenticity is insufficient based on news content only. It also needs to analyze social features of news. In this paper, we propose an approach to detect fake news on social media that covers both news content and social context. We use synonym-based feature extraction method and three different classifiers based on multidimensional dataset. Experimental result shows the effective as an accuracy way to define news authenticity on online news media.