Research map / Graph learning / Graph self-supervised learning
Label-invariant Augmentation for Semi-Supervised Graph Classification
Han Yue, Chunhui Zhang, Chuxu Zhang, Hongfu Liu
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 |
|---|---|---|---|
| remove negative pairs from the contrastive loss (positive-pair-only) (ablation) | GLA · four graph datasets, semi-supervised | accuracy | Improved |
| choose the hardest label-invariant augmented candidate instead of random or easiest (ablation) | GLA · four graph datasets, 50% labels | accuracy | Improved |
| filter augmentations that violate label invariance (GraphCL+Label-Invariant) (ablation) | GraphCL · four graph datasets, 50% labels | accuracy | Improved |
| augment in representation space rather than on raw graphs (ablation) | GLA · four graph datasets, 50% labels | accuracy | Improved |
| vary perturbation magnitude eta (ablation) | GLA · four graph datasets, 50% labels | accuracy | No clear effect |
| GLA vs JOAOv2 at 30% label ratio | JOAOv2 · average over 8 datasets, 30% labels | accuracy: 76.41 -> 76.25 | Hurt |
| GLA vs best baseline at 50% labels | JOAOv2 · average over 8 datasets, 50% labels | accuracy: 77.53 -> 77.85 | Improved |
| GLA vs best baseline at 70% labels | JOAOv2 · average over 8 datasets, 70% labels | accuracy: 78.26 -> 78.73 | Improved |
| GLA vs JOAOv2 on GITHUB | JOAOv2 · GITHUB, 50% labels | accuracy: 68.97 -> 66.59 | Hurt |
Closest papers by meaning
- Community-Invariant Graph Contrastive Learning (ICML 2024 · similarity 0.91)
- Graph Contrastive Learning Automated (ICML 2021 · similarity 0.90)
- Graph Contrastive Learning with Augmentations (NeurIPS 2020 · similarity 0.90)
- Perfect Alignment May be Poisonous to Graph Contrastive Learning (ICML 2024 · similarity 0.88)
- Edge Contrastive Learning: An Augmentation-Free Graph Contrastive Learning Model (AAAI 2025 · similarity 0.88)
- Hybrid-Collaborative Augmentation and Contrastive Sample Adaptive-Differential Awareness for Robust Attributed Graph Clustering (NeurIPS 2025 · 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.
