Fake News and Influential Figures, an American Love Story

Abstract

Fake news has been a pervasive topic for many years, affecting both political and social aspects of everyday life. Through the birth of various social media platforms, fake news has been allowed to spread at an even more rapid pace, as more people are exposed to it through sites such as Twitter and Facebook. Political figures often take advantage of the lack of fact-checking done on these sites, as they hope to gain as much favorable influence as possible for political power. This paper aims to shed light on the prevalence of fake news through identifying influential spreaders of fake news on Twitter. Donald Trump’s account is used to demonstrate fake news propagation through a social network. This is accomplished through formulating the influential users as a network, associating them with various centrality measures, and classifying their tweets. Techniques from natural language processing and machine learning are used to classify tweets. Linear Support Vector Machines, Naïve Bayes, and Logistic regression are used as methods of classification.