Modeling The Sharing and Diffusion Of Fake News in Social Media

TL;DR

This paper models the spread of fake news on social media by analyzing user features and applying machine learning to predict and identify likely fake news spreaders, aiming to curb misinformation dissemination.

Contribution

It introduces a comprehensive model combining user features, content style, and psychological factors to predict fake news spreaders using machine learning algorithms.

Findings

Influential users with more followers and engagement are more likely to spread fake news.

Machine learning models effectively predict potential fake news spreaders.

User connection strength and topical interests impact fake news dissemination.

Abstract

The use of social media platforms has been gradually increasing and fake news spreading is becoming an alarming issue nowadays. The spreading of fake news means disseminating false, confusing, and spurious information which hurts families, communities etc. As a result, this issue has to be resolved sooner so that we can limit the spread of fake news in the virtual world. One needs to identify the fake news spreader to address this issue. In this research, we have tried to reveal the users who are most likely to share fake news as well as the spread prediction that shared pieces of fake news in the social network. We take into account the users information, such as follower counts, like counts, and retweet counts along with users topical interests on different topics as well as connection strength by considering the follower-following ratio. We also consider the complexity features, stylistic features, and psychological effects of news. Finally, we applied different machine-learning algorithms to evaluate the performance of the proposed model. Our observation is that the probability of spreading a piece of news shared by users having more followers as well as more likes and retweet counts (aka influential users) is higher compared with other users.