Convolutional Set Matching for Graph Similarity

TL;DR

This paper presents GSimCNN, a convolutional neural network-based model for predicting graph similarity scores, achieving state-of-the-art results on real datasets despite the NP-hard nature of graph similarity computation.

Contribution

The paper introduces GSimCNN, a novel neural network approach for graph similarity prediction that effectively approximates complex graph distance metrics.

Findings

Achieves state-of-the-art performance on graph similarity search

Effective approximation of Graph Edit Distance using CNNs

Demonstrates scalability on real graph datasets

Abstract

We introduce GSimCNN (Graph Similarity Computation via Convolutional Neural Networks) for predicting the similarity score between two graphs. As the core operation of graph similarity search, pairwise graph similarity computation is a challenging problem due to the NP-hard nature of computing many graph distance/similarity metrics. We demonstrate our model using the Graph Edit Distance (GED) as the example metric. Experiments on three real graph datasets demonstrate that our model achieves the state-of-the-art performance on graph similarity search.