Curtailing Fake News Propagation with Psychographics

Abstract

Fake news is a widespread and concerning phenomenon. The goal of this article is twofold. First, we argue that curtailing fake news is better pursued by identifying its propagators than by classifying its content. Second, to validate our conceptual argument, we develop a system for curtailing the propagation of fake news on social media by identifying users who are susceptible to believing and propagating it. We anchor our approach in the debate about the ontological nature of truth, the empirical challenges of classifying fake news content, as well as the psychological and social origins of believing fake news. Through interpreting our model using modern explainable machine learning, we deepen our theoretical understanding of why people believe and share fake news, extend the applicability of our system beyond its original context, and provide guidelines for mitigating fake news propagation.