Research map / Graph learning / Graph self-supervised learning
Revisiting Positive Samples in Graph Contrastive Learning: From the Perspective of Message Passing
Lianze Shan, Ningchong Wang, Jitao Zhao, Di Jin, Dongxiao He
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 |
|---|---|---|---|
| add Energy Aware Propagation (ablation) | SPGCL · Cornell | accuracy (%): 50.27 -> 75.41 | Improved |
| add Energy Aware Propagation (ablation) | SPGCL · Actor | accuracy (%): 24.86 -> 37.23 | Improved |
| add Energy Aware Propagation (ablation) | SPGCL · Texas | accuracy (%): 68.91 -> 80.81 | Improved |
| add Energy Aware Propagation (ablation) | SPGCL · Cora | accuracy (%): 84.10 -> 85.44 | Improved |
| add Energy Aware Propagation (ablation) | SPGCL · CS | accuracy (%): 92.99 -> 95.03 | Improved |
| add Energy-guided Positive Sampling (ablation) | SPGCL · Cornell | accuracy (%): 66.76 -> 75.41 | Improved |
| add Energy-guided Positive Sampling (ablation) | SPGCL · Actor | accuracy (%): 35.00 -> 37.23 | Improved |
| add Energy-guided Positive Sampling (ablation) | SPGCL · Cora | accuracy (%): 84.91 -> 85.44 | Improved |
| add EPS on top of removed EAP (ablation) | SPGCL · Photo | accuracy (%): 93.21 -> 93.73 | Hurt |
| use SPGCL instead of StrGCL | contrastive learning · Cora | accuracy (%): 84.86 -> 85.44 | Improved |
| use SPGCL instead of GraphACL | contrastive learning · Cornell | accuracy (%): 60.54 -> 75.41 | Improved |
| use SPGCL instead of BGRL | contrastive learning · Photo clustering | NMI: 0.6841 -> 0.7135 | Improved |
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 0.92)
- Improving Graph Contrastive Learning via Adaptive Positive Sampling (CVPR 2024 · similarity 0.92)
- 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)
- Perfect Alignment May be Poisonous to Graph Contrastive Learning (ICML 2024 · similarity 0.89)
- A New Mechanism for Eliminating Implicit Conflict in Graph Contrastive Learning (AAAI 2024 · 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.
