Learning Graph Edit Distance by Graph Neural Networks
This work addresses graph similarity computation, which is important for applications like keyword spotting, but it is incremental as it builds on existing deep learning and graph edit distance methods.
The paper tackles the problem of computing graph edit distance by proposing a new framework that combines deep metric learning with traditional approximations, using a message passing neural network. The method shows superior performance in graph retrieval for handwritten words and competitive results on a graph similarity benchmark.
The emergence of geometric deep learning as a novel framework to deal with graph-based representations has faded away traditional approaches in favor of completely new methodologies. In this paper, we propose a new framework able to combine the advances on deep metric learning with traditional approximations of the graph edit distance. Hence, we propose an efficient graph distance based on the novel field of geometric deep learning. Our method employs a message passing neural network to capture the graph structure, and thus, leveraging this information for its use on a distance computation. The performance of the proposed graph distance is validated on two different scenarios. On the one hand, in a graph retrieval of handwritten words~\ie~keyword spotting, showing its superior performance when compared with (approximate) graph edit distance benchmarks. On the other hand, demonstrating competitive results for graph similarity learning when compared with the current state-of-the-art on a recent benchmark dataset.