Research map / Graph learning / Graph algorithms and clustering
D2Match: Leveraging Deep Learning and Degeneracy for Subgraph Matching
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 |
|---|---|---|---|
| Replace GNN-embedding-only matching with D2Match's subtree-indicator + GNN design (subtree matching via tree aggregation) | subgraph matching on uniformly distributed data (starred datasets) · Proteins*/Aids* etc. vs SimGNN | accuracy: e.g., Proteins* 83.4 vs 64.6 for SimGNN | Helped |
| Use D2Match (subtree-matching GNN) instead of baselines on original benchmarks | subgraph matching accuracy on 7 real and synthetic datasets · Synthetic | accuracy: 74.3 vs 70.5 (SimGNN) | Helped |
| Compare with exact CO solvers on runtime | D2Match vs GQL/QSI/CECI/LFTJ · Proteins, Enzymes, Aids | runtime (s): D2Match 1.12/1.18/1.88 vs best CO 3.82/2.98/6.94 | Helped |
| Compare with exact CO solvers on Mutag runtime | D2Match vs QSI · Mutag | runtime (s): D2Match 1.04 vs QSI 0.16 | Hurt |
| Scale to large graphs versus CO solvers | D2Match on DD and Firstmm · DD and Firstmm | success rate (%): DD 99.2 (GQL 100.0); Firstmm 60.0 (LFTJ 77.2) | Mixed |
| Remove GNN module (subtree-only) on biased vs unbiased synthetic data (ablation) | D2Match · Synthetic and Synthetic+ | accuracy | No clear effect |
| Add chordless cycle supernodes (ablation) | D2Match · Synthetic | accuracy | Helped |
Closest papers by meaning
- Charting the Design Space of Neural Graph Representations for Subgraph Matching (ICLR 2025 · similarity 0.88)
- Discrete Cycle-Consistency Based Unsupervised Deep Graph Matching (AAAI 2024 · similarity 0.88)
- Iteratively Refined Early Interaction Alignment for Subgraph Matching based Graph Retrieval (NeurIPS 2024 · similarity 0.88)
- Learning Partial Graph Matching via Optimal Partial Transport (ICLR 2025 · similarity 0.88)
- Position: Graph Matching Systems Deserve Better Benchmarks (ICML 2025 · similarity 0.87)
- Translating Subgraphs to Nodes Makes Simple GNNs Strong and Efficient for Subgraph Representation Learning (ICML 2024 · similarity 0.87)
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.