Higher-Order Explanations of Graph Neural Networks via Relevant Walks

TL;DR

This paper introduces GNN-LRP, a method for explaining graph neural networks by identifying relevant walks in the input graph, making GNN decisions more interpretable across various applications.

Contribution

The paper proposes a novel higher-order explanation technique for GNNs using nested attribution, applicable to multiple GNN architectures and diverse domains.

Findings

GNN-LRP effectively identifies relevant walks in graphs.

The method provides insights into GNN decisions in text, chemistry, and image tasks.

Applicable to a broad range of GNN models.

Abstract

Graph Neural Networks (GNNs) are a popular approach for predicting graph structured data. As GNNs tightly entangle the input graph into the neural network structure, common explainable AI approaches are not applicable. To a large extent, GNNs have remained black-boxes for the user so far. In this paper, we show that GNNs can in fact be naturally explained using higher-order expansions, i.e. by identifying groups of edges that jointly contribute to the prediction. Practically, we find that such explanations can be extracted using a nested attribution scheme, where existing techniques such as layer-wise relevance propagation (LRP) can be applied at each step. The output is a collection of walks into the input graph that are relevant for the prediction. Our novel explanation method, which we denote by GNN-LRP, is applicable to a broad range of graph neural networks and lets us extract practically relevant insights on sentiment analysis of text data, structure-property relationships in quantum chemistry, and image classification.