Research map / Graph learning / GNN architectures
Learning on Attribute-Missing Graphs
Xu Chen, Siheng Chen, Jiangchao Yao, Huangjie Zheng, Ya Zhang, Ivor Wai-Hung Tsang
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 cross-reconstruction terms (ablation) | SAT(GCN) · Cora node classification (X only) | accuracy: 0.3036 -> 0.7644 | Improved |
| add self-reconstruction terms (ablation) | SAT(GCN) · Cora node classification (X only) | accuracy: 0.7074 -> 0.7644 | Improved |
| add adversarial distribution matching (ablation) | SAT(GCN) · Cora node classification (A+X) | accuracy: 0.8231 -> 0.8327 | Improved |
| add adversarial distribution matching (ablation) | SAT(GCN) · Citeseer node classification (A+X) | accuracy: 0.6609 -> 0.6599 | No clear effect |
| add cross-reconstruction terms (ablation) | SAT(GCN) · Cora node classification (A+X) | accuracy: 0.3402 -> 0.8327 | Improved |
| add cross-reconstruction terms (ablation) | SAT(GCN) · Steam profiling | Recall@10: 0.4611 -> 0.4933 | Improved |
| add adversarial distribution matching (ablation) | SAT(GCN) · Cora profiling | Recall@10: 0.1356 -> 0.1508 | Improved |
| add cross-reconstruction terms (ablation) | SAT(GCN) · Cora link prediction (A+X) | AUC: 0.8048 -> 0.8554 | Improved |
| add self-reconstruction terms (ablation) | SAT(GCN) · Coauthor-CS link prediction (A+X) | AUC: 0.8166 -> 0.9077 | Improved |
| add adversarial distribution matching (ablation) | SAT(GCN) · Coauthor-CS link prediction (A+X) | AUC: 0.9005 -> 0.9077 | Improved |
| use GAT instead of GCN backbone (ablation) | SAT · Cora node classification (X only) | accuracy: 0.7644 -> 0.7937 | Improved |
| feed SAT(GCN)-restored attributes to GCN classifier | GCN classifier · Cora node classification | accuracy: 0.7631 -> 0.8327 | Improved |
Closest papers by meaning
- Wasserstein Graph Neural Networks for Graphs With Missing Attributes (TPAMI 2025 · similarity 0.89)
- Attribute-Missing Graph Clustering Network (AAAI 2024 · similarity 0.86)
- Zero-Shot Generalization of GNNs over Distinct Attribute Domains (ICML 2025 · similarity 0.86)
- Graph Neural Networks with Soft Association between Topology and Attribute (AAAI 2024 · similarity 0.86)
- Propagate and Inject: Revisiting Propagation-Based Feature Imputation for Graphs with Partially Observed Features (ICML 2025 · similarity 0.85)
- Generalization, Expressivity, and Universality of Graph Neural Networks on Attributed Graphs (ICLR 2025 · similarity 0.85)
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.
