Graph Neural Network Causal Explanation via Neural Causal Models

TL;DR

This paper introduces { ame}, a causal inference-based GNN explainer that identifies truly causally relevant subgraphs, overcoming limitations of association-based methods and improving explanation accuracy.

Contribution

It develops a novel GNN causal explainer using neural causal models to accurately identify causally relevant subgraphs for GNN predictions.

Findings

Outperforms existing explainers on synthetic graphs

Achieves higher accuracy in real-world graph explanations

Effectively uncovers causally relevant subgraphs

Abstract

Graph neural network (GNN) explainers identify the important subgraph that ensures the prediction for a given graph. Until now, almost all GNN explainers are based on association, which is prone to spurious correlations. We propose {\name}, a GNN causal explainer via causal inference. Our explainer is based on the observation that a graph often consists of a causal underlying subgraph. {\name} includes three main steps: 1) It builds causal structure and the corresponding structural causal model (SCM) for a graph, which enables the cause-effect calculation among nodes. 2) Directly calculating the cause-effect in real-world graphs is computationally challenging. It is then enlightened by the recent neural causal model (NCM), a special type of SCM that is trainable, and design customized NCMs for GNNs. By training these GNN NCMs, the cause-effect can be easily calculated. 3) It uncovers the subgraph that causally explains the GNN predictions via the optimized GNN-NCMs. Evaluation results on multiple synthetic and real-world graphs validate that {\name} significantly outperforms existing GNN explainers in exact groundtruth explanation identification