Summary of Graph Neural Backdoor: Fundamentals, Methodologies, Applications, and Future Directions, by Xiao Yang et al.
Graph Neural Backdoor: Fundamentals, Methodologies, Applications, and Future Directions
by Xiao Yang, Gaolei Li, Jianhua Li
First submitted to arxiv on: 15 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
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 Graph Neural Networks (GNNs) have revolutionized various applications, including recommender systems, molecular structure prediction, and social media analysis, despite their potential vulnerability to backdoor attacks. Researchers have empirically demonstrated that GNNs can be poisoned by triggers, leading to malicious outputs. The lack of comprehensive investigation into this field has prompted the need for a dedicated survey on GNN backdoors. This paper proposes such a survey, outlining the fundamental definition of GNNs, summarizing and categorizing current attacks and defenses, analyzing their applicability and use cases, and exploring potential research directions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary GNNs are machine learning models that work with graph data. They’re very good at predicting things like what products you might buy or how molecules behave. But recently, scientists discovered that these models can be tricked into making bad predictions by adding special “triggers” to the data they learn from. This is a big problem because it could let hackers control what GNNs do. To understand this better, researchers are writing a report that explains all the different types of attacks and defenses people have found so far. |
Keywords
* Artificial intelligence * Gnn * Machine learning
