Research map / Graph learning / Graph algorithms and clustering
Graph Matching Networks for Learning the Similarity of Graph Structured Objects
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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-graph attention-based matching to graph message passing | GNN graph similarity model · Synthetic graph edit-distance learning | pair AUC: 88.8 -> 95.0; 92.1 -> 96.6; 95.9 -> 97.4; 88.5 -> 93.8 | Helped |
| Add cross-graph attention-based matching (ablation) | GNN graph similarity model · Function Similarity Search | pair AUC: 97.71 -> 99.28 | Helped |
| Add cross-graph attention-based matching (ablation) | GNN graph similarity model · COIL-DEL | pair AUC: 98.58 -> 98.97 | Helped |
| Use Siamese fusion of graph vectors instead of a prespecified similarity metric (ablation) | GNN graph embedding model · Function Similarity Search | pair AUC: 97.71 -> 97.76 | No clear effect |
| Use Siamese fusion of graph vectors instead of a prespecified similarity metric (ablation) | GNN graph embedding model · COIL-DEL | pair AUC: 98.58 -> 98.76 | Helped |
| Increase the number of propagation steps (ablation) | GNN and GMN · Synthetic graph edit-distance learning | pair AUC | Helped |
| Increase the number of propagation steps (ablation) | GNN and GMN · Function Similarity Search | pair AUC | Helped |
| Use 0 propagation steps, removing graph-structure propagation (ablation) | GNN or GMN · Function Similarity Search | pair AUC | Hurt |
| Replace GNN message passing with GCN graph convolutions (ablation) | Graph similarity model · Function Similarity Search | pair AUC: 97.71 -> 96.67 | Hurt |
| Replace GNN message passing with GCN graph convolutions (ablation) | Graph similarity model · COIL-DEL | pair AUC: 98.58 -> 94.80 | Hurt |
| Use pair training instead of triplet training (ablation) | GSL model · Synthetic graph edit-distance learning | performance | Helped |
| Use triplet training instead of pair training (ablation) | GSL model · Function Similarity Search | performance | Helped |
Closest papers by meaning
- Position: Graph Matching Systems Deserve Better Benchmarks (ICML 2025 · similarity 0.87)
- Charting the Design Space of Neural Graph Representations for Subgraph Matching (ICLR 2025 · similarity 0.87)
- Iteratively Refined Early Interaction Alignment for Subgraph Matching based Graph Retrieval (NeurIPS 2024 · similarity 0.87)
- Exploring Consistency in Graph Representations: from Graph Kernels to Graph Neural Networks (NeurIPS 2024 · similarity 0.87)
- RAGraph: A General Retrieval-Augmented Graph Learning Framework (NeurIPS 2024 · similarity 0.86)
- Graph Segmentation and Contrastive Enhanced Explainer for Graph Neural Networks (AAAI 2025 · similarity 0.86)
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.