Computing Graph Edit Distance via Neural Graph Matching
Summary: Introduce GEDGNN, a GNN that jointly predicts GED and a node-matching matrix, enabling recovery of edit paths via a k-best matching post-processing step—bridging regression-based GNNs and constructive edit-path algorithms. Achieves 4.9–74.3% MAE reduction vs prior GNNs and 53.6–88.1% MAE reduction on edit-path-derived GED vs state-of-the-art Noah across real and synthetic graphs. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Chengzhi Piao
- 2. Tingyang Xu
- 3. Xiangguo Sun
- 4. Yu Rong
- 5. Kangfei Zhao
- 6. Hong Cheng
Incoming Citations (Sorted by Pagerank)
Showing 9 of 9 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 4 of 4 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 3,001 | Neural Subgraph Counting with Wasserstein Estimator | 2022 | SIGMOD | 7.7404487e-05 |
| 3,778 | A Learned Sketch for Subgraph Counting | 2021 | SIGMOD | 6.7747398e-05 |
| 3,849 | TaGSim: Type-aware Graph Similarity Learning and Computation | 2022 | VLDB | 6.7064042e-05 |
| 4,837 | Entity Resolution with Hierarchical Graph Attention Networks | 2022 | SIGMOD | 5.8892326e-05 |
Previous
Page 1 / 1
Next