USER: Unsupervised Structural Entropy-based Robust Graph Neural Network
arXiv:2302.05889 · doi:10.1609/aaai.v37i8.26219
Abstract
Unsupervised/self-supervised graph neural networks (GNN) are vulnerable to inherent randomness in the input graph data which greatly affects the performance of the model in downstream tasks. In this paper, we alleviate the interference of graph randomness and learn appropriate representations of nodes without label information. To this end, we propose USER, an unsupervised robust version of graph neural networks that is based on structural entropy. We analyze the property of intrinsic connectivity and define intrinsic connectivity graph. We also identify the rank of the adjacency matrix as a crucial factor in revealing a graph that provides the same embeddings as the intrinsic connectivity graph. We then introduce structural entropy in the objective function to capture such a graph. Extensive experiments conducted on clustering and link prediction tasks under random-noises and meta-attack over three datasets show USER outperforms benchmarks and is robust to heavier randomness.
References in corpus (17)
- Community detection in graphs
- DeepWalk: Online Learning of Social Representations
- Semi-Supervised Classification with Graph Convolutional Networks
- Inductive Representation Learning on Large Graphs
- Learning deep representations by mutual information estimation and maximization
- Simplifying Graph Convolutional Networks
- Graph Contrastive Learning with Adaptive Augmentation
- Entropy measures for complex networks: Toward an information theory of complex topologies
- Learning Graph Embedding with Adversarial Training Methods
- Adversarial Examples on Graph Data: Deep Insights into Attack and Defense
- A Survey of Adversarial Learning on Graphs
- Graph Structure Learning for Robust Graph Neural Networks
- Attributed Graph Clustering: A Deep Attentional Embedding Approach
- Graph InfoClust: Leveraging cluster-level node information for unsupervised graph representation learning
- Adversarial Attacks and Defenses on Graphs: A Review, A Tool and Empirical Studies
- Attributed Graph Clustering via Adaptive Graph Convolution
- Learning to Drop: Robust Graph Neural Network via Topological Denoising