Automatically Identifying Fake News in Popular Twitter Threads

Abstract

Information quality in social media is an increasingly important issue, but web-scale data hinders experts' ability to assess and correct much of the inaccurate content, or "fake news," present in these platforms. This paper develops a method for automating fake news detection on Twitter by learning to predict accuracy assessments in two credibility-focused Twitter datasets: CREDBANK, a crowdsourced dataset of accuracy assessments for events in Twitter, and PHEME, a dataset of potential rumors in Twitter and journalistic assessments of their accuracies. We apply this method to Twitter content sourced from BuzzFeed's fake news dataset and show models trained against crowdsourced workers outperform models based on journalists' assessment and models trained on a pooled dataset of both crowdsourced workers and journalists. All three datasets, aligned into a uniform format, are also publicly available. A feature analysis then identifies features that are most predictive for crowdsourced and journalistic accuracy assessments, results of which are consistent with prior work. We close with a discussion contrasting accuracy and credibility and why models of nonexperts outperform models of journalists for fake news detection in Twitter.

Analysis of Information Spreading by Social Media Based on Emotion and Empathy

“…Thus, if we can calculate the reliability of the content posted online based on the feedback on social media, we might be able to protect innocent users from malicious ones. Various recent studies have focused on rumor detection or judgment of fake news on social media [22][23][24][25][26][27][28]. Most of these studies analyzed the contents or behaviors of the users or estimated the reliability of their remarks.…”

Facebook fake profile identification: technical and ethical considerations

“…in the process of detecting fake news. In [3], authors claim to have developed a process for automating fake news detection on Twitter. They achieved it by learning how to forecast precision assessments in two dedicated Twitter datasets.…”

Can Machines Learn to Detect Fake News? A Survey Focused on Social Media

“…The APIs usually provides the content of the platform in structured data or plain text, thus reducing the preprocessing step that is commonly used with web crawlers used to filter the information of interest from web pages. [22][16] [23] Another reason for this is that most of newspapers are just too serious and express more a generic political opinion compared to the social networks that express individual opinions of many different users with different beliefs, contexts and cultural backgrounds. Also it is very difficult to find an expressive newspaper that diffuse rumours and fake news, as the assurance of information quality is part of the newspaper's main process.…”