Detecting Fake News on Social Media

Abstract

As Web 2.0 technologies have turned the Internet into an interactive medium, users dominate the field. With the spread of social media, the Internet has become much more user-oriented. In contrast to traditional media, social media's lack of control mechanisms makes the accuracy of spreading news questionable. This brings us to the significance of fact-checking platforms. This study investigates the antecedents of spreading false news in Turkey. The purpose of the study is to determine the features of fake news. For this purpose, teyit.org, the biggest fact-checking platform in Turkey, has been chosen for analysis. The current study shows fake news to be detectable based on four features: Propagation, User Type, Social Media Type, and Formatting. According to the logistic regression analysis, the study's model obtained 86.7% accuracy. The study demonstrates that Facebook increases the likelihood of news being fake compared to Twitter or Instagram. Emoji usage is also statistically significant in terms of increasing the probability of fake news. Unexpectedly, the impact of photos or videos was found statistically insignificant.