# Fake News Detection with Deep Diffusive Network Model

**Authors:** Jiawei Zhang, Limeng Cui, Yanjie Fu, Fisher B. Gouza  
**arXiv:** [1805.08751](https://arxiv.org/abs/1805.08751)  
**Date:** 2018-05-22  
**Summary page:** https://www.summarizepaper.com/en/paper/1805.08751v1/

## Key points

- Proliferation of online social networks has led to a surge in dissemination of fake news for commercial and political purposes
- Fake news has significant repercussions on offline society as unsuspecting users can easily fall prey to deceptive information
- Researchers Jiawei Zhang, Limeng Cui, Yanjie Fu, and Fisher B. Gouza developed methodologies and algorithms for detecting fake news articles, creators, and subjects within online social networks
- Investigation addresses challenges posed by the elusive nature of fake news and intricate connections among news articles, creators, and subjects
- Introduce an automatic fake news credibility inference model named FakeDetector that leverages explicit and latent features extracted from textual information
- FakeDetector employs a deep diffusive network model to learn representations of news articles, creators, and subjects simultaneously
- Extensive experiments using real-world dataset showed the effectiveness of FakeDetector in identifying fake news within online social networks
- Research contributes significantly to combating the spread of misinformation in digital spaces through advanced detection techniques

## Summary

In recent years, the proliferation of online social networks has led to a surge in the dissemination of fake news for various commercial and political purposes. This phenomenon has had significant repercussions on offline society as unsuspecting users can easily fall prey to deceptive information. Recognizing the urgency of enhancing the trustworthiness of information shared on online platforms, researchers Jiawei Zhang, Limeng Cui, Yanjie Fu, and Fisher B. Gouza have delved into the development of methodologies and algorithms for detecting fake news articles, creators, and subjects within online social networks. The team's investigation addresses the complex challenges posed by the elusive nature of fake news and intricate connections among news articles, creators, and subjects. Through meticulous data analysis, they introduce an innovative automatic fake news credibility inference model named FakeDetector. Leveraging a combination of explicit and latent features extracted from textual information, FakeDetector employs a deep diffusive network model to simultaneously learn representations of news articles, creators, and subjects. To evaluate the efficacy of their proposed model, extensive experiments were conducted using a real-world dataset comprising fake news instances. Comparative analyses with several state-of-the-art models showcased the effectiveness of FakeDetector in accurately identifying fake news within online social networks. By shedding light on advanced detection techniques and showcasing their practical application through empirical studies,this research contributes significantly to ongoing efforts aimed at combating the spread of misinformation in digital spaces.