Research map / Graph learning / Graph self-supervised learning
Uncovering the Structural Fairness in Graph Contrastive Learning
Ruijia Wang, Xiao Wang, Chuan Shi, Le Song
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use GRADE instead of GCN | node classification models · Cora | Bias: 1.70 -> 0.48 | Improved |
| Use GRADE instead of GCN | node classification models · Citeseer | Bias: 1.85 -> 1.52 | Improved |
| Use GRADE instead of GCN | node classification models · Photo | Bias: 0.28 -> 0.20 | Improved |
| Use GRADE instead of GCN | node classification models · Computer | Bias: 0.50 -> 0.35 | Improved |
| Add random interpolation when augmenting tail-node neighborhoods (ablation) | GRADE · Cora | G. Mean: 90.86 -> 92.87 | Improved |
| Add random interpolation when augmenting tail-node neighborhoods (ablation) | GRADE · Citeseer | G. Mean: 84.96 -> 85.88 | Improved |
| Add random deleting when purifying head-node neighborhoods (ablation) | GRADE · Cora | G. Mean: 90.54 -> 92.87 | Improved |
| Add random deleting when purifying head-node neighborhoods (ablation) | GRADE · Citeseer | G. Mean: 84.21 -> 85.88 | Improved |
| Add random interpolation and random deleting (ablation) | GRADE · Cora | G. Mean: 89.97 -> 92.87 | Improved |
| Add random interpolation and random deleting (ablation) | GRADE · Citeseer | G. Mean: 83.00 -> 85.88 | Improved |
| Choose a suitable tail/head threshold rather than an overly small or large threshold | GRADE · Cora | Micro-F1 | Mixed |
| Increase feature drop rate too far | GRADE graph augmentation · Cora | Micro-F1 | Hurt |
Closest papers by meaning
- Does GCL Need a Large Number of Negative Samples? Enhancing Graph Contrastive Learning with Effective and Efficient Negative Sampling (AAAI 2025 · similarity 0.91)
- ReGCL: Rethinking Message Passing in Graph Contrastive Learning (AAAI 2024 · similarity 0.91)
- S3GCL: Spectral, Swift, Spatial Graph Contrastive Learning (ICML 2024 · similarity 0.90)
- Beyond Homophily: Graph Contrastive Learning with Macro-Micro Message Passing (AAAI 2025 · similarity 0.90)
- Community-Invariant Graph Contrastive Learning (ICML 2024 · similarity 0.90)
- Fair Graph U-Net: A Fair Graph Learning Framework Integrating Group and Individual Awareness (AAAI 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
