Reproduce “Some Like it Hoax: Automated Fake News Detection in Social Networks” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Eugenio Tacchini, Gabriele Ballarin, Marco L. Della Vedova, Stefano Moret, Luca de Alfaro
An investigation into automated fake news detection found that user interaction patterns, specifically Facebook "likes", offer a highly accurate signal for identifying misinformation. An adapted harmonic boolean crowdsourcing algorithm achieved over 99% accuracy using less than 1% of posts as training data and demonstrated strong transferability to unseen communities.
Key insights
- User interaction patterns, particularly "likes," serve as a highly predictive and robust signal for determining the veracity of social media posts.
- The adapted Harmonic Boolean Label Crowdsourcing algorithm is exceptionally efficient, requiring minimal labeled training data to achieve high accuracy and demonstrating strong generalization capabilities across diverse user communities.
- While user behavior shows polarization, the methods are robust enough to classify posts accurately even when a significant portion of users engages with both reliable and unreliable content.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.