Research map / Graph learning / Graph self-supervised learning
Analyzing Data-Centric Properties for Graph Contrastive Learning
Puja Trivedi, Ekdeep Singh Lubana, Mark Heimann, Danai Koutra, Jayaraman J. Thiagarajan
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 |
|---|---|---|---|
| train graph SSL methods with generic graph augmentations vs untrained encoder | GraphCL, GAE, AAGAE, SpecCL, SimSiam, BYOL · 7 graph benchmarks (Reddit etc.) | kNN accuracy | No clear effect |
| increase invariance to generic augmentations | GraphCL and other graph SSL methods · Reddit | kNN accuracy | No clear effect |
| use content-aware edge dropping instead of random edge dropping | GraphCL · Synthetic, increasing style ratio | linear probe accuracy | Improved |
| use content-aware edge dropping instead of random edge dropping | SpecLoss · Synthetic motif+tree graph dataset | linear probe accuracy | Improved |
| use content-aware augmentations | GAE and AAGAE reconstruction methods · Synthetic motif+tree graph dataset | linear probe accuracy | No clear effect |
| use learned augmentor (AD-GCL) | GraphCL-style contrastive learning · Synthetic motif+tree graph dataset | linear probe accuracy | Hurt |
| add content-preserving edge dropping to JOAO augmentation prior | JOAO · Synthetic motif+tree graph dataset | linear probe accuracy | No clear effect |
| use implicit weight-space perturbation (SimGRACE) | SimGRACE · Synthetic motif+tree graph dataset | linear probe accuracy | Hurt |
| use content-preserving instead of random edge dropping at 20% strength | GraphCL · Synthetic, kappa=6 | separability | Improved |
| use content-preserving instead of random edge dropping at 20% strength | GraphCL · Synthetic, kappa=6 | invariance | No clear effect |
| raise generic augmentation strength to 60% | GraphCL · Synthetic, kappa=6 | invariance and separability | Hurt |
| use biochemistry-inspired content-aware augmentations instead of mask edge features | GraphCL-style SSL · BACE | invariance and separability | Improved |
Closest papers by meaning
- An Empirical Study of Graph Contrastive Learning (NeurIPS 2021 · similarity 0.92)
- Graph Contrastive Learning Automated (ICML 2021 · similarity 0.91)
- Perfect Alignment May be Poisonous to Graph Contrastive Learning (ICML 2024 · similarity 0.91)
- S3GCL: Spectral, Swift, Spatial Graph Contrastive Learning (ICML 2024 · similarity 0.89)
- Community-Invariant Graph Contrastive Learning (ICML 2024 · similarity 0.89)
- Unified Graph Augmentations for Generalized Contrastive Learning on Graphs (NeurIPS 2024 · similarity 0.88)
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.
