A Gradient-Based Explanation Method for Node Classification Using Graph Convolutional Networks

Abstract

Explainable artificial intelligence is a method that explains how a complex model (e.g., a deep neural network)yields its output from a given input. Recently, graph-type data have been widely used in various fields, anddiverse graph neural networks (GNNs) have been developed for graph-type data. However, methods to explainthe behavior of GNNs have not been studied much, and only a limited understanding of GNNs is currentlyavailable. Therefore, in this paper, we propose an explanation method for node classification using graphconvolutional networks (GCNs), which is a representative type of GNN. The proposed method finds out whichfeatures of each node have the greatest influence on the classification of that node using GCN. The proposedmethod identifies influential features by backtracking the layers of the GCN from the output layer to the inputlayer using the gradients. The experimental results on both synthetic and real datasets demonstrate that theproposed explanation method accurately identifies the features of each node that have the greatest influence onits classification.