Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SEGMN: A Structure-Enhanced Graph Matching Network for Graph Similarity Learning

Wenjun Wang, Jiacheng Lu, Kejia Chen, Zheng Liu, Shilong Sang

TL;DR

Equipped with a dual embedding learning module and a structure perception matching module, SEGMN achieves structure enhancement in both embedding learning and cross-graph matching, and Experimental results on benchmark datasets demonstrate that SEGMN outperforms the state-of-the-art GSC methods in the GED regression task.

Abstract

Graph similarity computation (GSC) aims to quantify the similarity score between two graphs. Although recent GSC methods based on graph neural networks (GNNs) take advantage of intra-graph structures in message passing, few of them fully utilize the structures presented by edges to boost the representation of their connected nodes. Moreover, previous cross-graph node embedding matching lacks the perception of the overall structure of the graph pair, due to the fact that the node representations from GNNs are confined to the intra-graph structure, causing the unreasonable similarity score. Intuitively, the cross-graph structure represented in the assignment graph is helpful to rectify the inappropriate matching. Therefore, we propose a structure-enhanced graph matching network (SEGMN). Equipped with a dual embedding learning module and a structure perception matching module, SEGMN achieves structure enhancement in both embedding learning and cross-graph matching. The dual embedding learning module incorporates adjacent edge representation into each node to achieve a structure-enhanced representation. The structure perception matching module achieves cross-graph structure enhancement through assignment graph convolution. The similarity score of each cross-graph node pair can be rectified by aggregating messages from structurally relevant node pairs. Experimental results on benchmark datasets demonstrate that SEGMN outperforms the state-of-the-art GSC methods in the GED regression task, and the structure perception matching module is plug-and-play, which can further improve the performance of the baselines by up to 25%.

SEGMN: A Structure-Enhanced Graph Matching Network for Graph Similarity Learning

TL;DR

Equipped with a dual embedding learning module and a structure perception matching module, SEGMN achieves structure enhancement in both embedding learning and cross-graph matching, and Experimental results on benchmark datasets demonstrate that SEGMN outperforms the state-of-the-art GSC methods in the GED regression task.

Abstract

Graph similarity computation (GSC) aims to quantify the similarity score between two graphs. Although recent GSC methods based on graph neural networks (GNNs) take advantage of intra-graph structures in message passing, few of them fully utilize the structures presented by edges to boost the representation of their connected nodes. Moreover, previous cross-graph node embedding matching lacks the perception of the overall structure of the graph pair, due to the fact that the node representations from GNNs are confined to the intra-graph structure, causing the unreasonable similarity score. Intuitively, the cross-graph structure represented in the assignment graph is helpful to rectify the inappropriate matching. Therefore, we propose a structure-enhanced graph matching network (SEGMN). Equipped with a dual embedding learning module and a structure perception matching module, SEGMN achieves structure enhancement in both embedding learning and cross-graph matching. The dual embedding learning module incorporates adjacent edge representation into each node to achieve a structure-enhanced representation. The structure perception matching module achieves cross-graph structure enhancement through assignment graph convolution. The similarity score of each cross-graph node pair can be rectified by aggregating messages from structurally relevant node pairs. Experimental results on benchmark datasets demonstrate that SEGMN outperforms the state-of-the-art GSC methods in the GED regression task, and the structure perception matching module is plug-and-play, which can further improve the performance of the baselines by up to 25%.

Paper Structure

This paper contains 21 sections, 15 equations, 4 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Preliminaries
  4. Method
  5. Overview
  6. Dual embedding learning
  7. Cross-graph interaction
  8. Structure perception matching
  9. Similarity matrix learning
  10. Loss function
  11. Experiment
  12. Experimental settings
  13. Overall performance (RQ1)
  14. Ablation study (RQ2)
  15. Portability and case study of the structure perception matching module (RQ3&RQ4)
  16. ...and 6 more sections

Figures (4)

  • Figure 1: An example of cross-graph structural correction. The figure shows a pair of graphs with their partial assignment graph and node embedding space. The closer the two embeddings are in the space, the more similar they are supposed to be. $S_{1a}$, $S_{3c}$ represent the similarity scores of node pair $(v^1, v^a)$ and $(v^3, v^c)$.
  • Figure 2: Framework of SEGMN.
  • Figure 3: Relationship between structure perception matching and structure-enhanced algorithm.
  • Figure 4: Case study. The dotted line in $G_2$ shows the graph edit path, and if the edge is connected, they are the same. Similarity matrix $S_1$ is the similarity matrix from cross-graph interaction and $S'_1$ is the matrix after structure perception matching.