Summary of Rumor Detection with a Novel Graph Neural Network Approach, by Tianrui Liu et al.
Rumor Detection with a novel graph neural network approach
by Tianrui Liu, Qi Cai, Changxin Xu, Bo Hong, Fanghao Ni, Yuxin Qiao, Tsungwei Yang
First submitted to arxiv on: 24 Mar 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed model jointly learns representations of user correlation and information propagation to detect rumors on social media. It leverages graph neural networks to learn user correlation from a bipartite graph describing correlations between users and source tweets, and represents information propagation with a tree structure. The learned representations are combined to classify rumors. To analyze the cost of adversarial attacks, the model develops a greedy attack scheme for three types: graph, comment, and joint attacks. Evaluation on two public datasets shows the proposed model outperforms state-of-the-art models in rumor detection, particularly for early detection. It also demonstrates robustness against adversarial attacks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Rumors spread quickly on social media, causing panic and fear. To stop this, researchers need to figure out how to detect rumors as soon as possible. Most current methods focus on how information spreads between people. However, few look at how users might work together to make a rumor popular. This paper proposes a new way to detect rumors by learning about both user connections and information spread. It uses special neural networks called graph neural networks to understand user relationships from a special kind of graph. Then, it combines this with another network that learns about how information spreads. To keep the model safe from bad actors, the researchers developed a plan to analyze three types of attacks: one that targets the graph, one that targets comments, and one that targets both. The results show that this new method does better than other methods at detecting rumors and is more resistant to attacks. |
