Banner Banner

Relevant Walk Search for Explaining Graph Neural Networks

Ping Xiong
Thomas Schnake
Michael Gastegger
Grégoire Montavon
Klaus Robert Müller
Shinichi Nakajima

April 24, 2023

Graph Neural Networks (GNNs) have become important machine learning tools for graph analysis, and its explainability is crucial for safety, fairness, and robustness. Layer-wise relevance propagation for GNNs (GNN-LRP) evaluates the relevance of walks to reveal important information flows in the network, and provides higher-order explanations, which have been shown to be superior to the lower-order, i.e., node-/edge-level, explanations. However, identifying relevant walks by GNN-LRP requires exponential computational complexity with respect to the network depth, which we will remedy in this paper. Specifically, we propose polynomial-time algorithms for finding top-� relevant walks, which drastically reduces the computation and thus increases the applicability of GNN-LRP to large-scale problems. Our proposed algorithms are based on the max-product algorithm---a common tool for finding the maximum likelihood configurations in probabilistic graphical models---and can find the most relevant walks exactly at the neuron level and approximately at the node level. Our experiments demonstrate the performance of our algorithms at scale and their utility across application domains, i.e., on epidemiology, molecular, and natural language benchmarks. We provide our codes under github.com/xiong-ping/rel_walk_gnnlrp.