Robust Counterfactual Explanations on Graph Neural Networks

TL;DR

This paper introduces a new method for generating robust counterfactual explanations for Graph Neural Networks that are resistant to noise and align with human intuition by modeling common decision boundaries across similar graphs.

Contribution

The paper proposes a novel approach that explicitly models GNN decision logic on similar graphs to produce explanations robust to noise and more intuitive.

Findings

Outperforms existing explanation methods on multiple datasets

Produces explanations that are resistant to input noise

Aligns well with human intuition by affecting predictions when edges are removed

Abstract

Massive deployment of Graph Neural Networks (GNNs) in high-stake applications generates a strong demand for explanations that are robust to noise and align well with human intuition. Most existing methods generate explanations by identifying a subgraph of an input graph that has a strong correlation with the prediction. These explanations are not robust to noise because independently optimizing the correlation for a single input can easily overfit noise. Moreover, they do not align well with human intuition because removing an identified subgraph from an input graph does not necessarily change the prediction result. In this paper, we propose a novel method to generate robust counterfactual explanations on GNNs by explicitly modelling the common decision logic of GNNs on similar input graphs. Our explanations are naturally robust to noise because they are produced from the common decision boundaries of a GNN that govern the predictions of many similar input graphs. The explanations also align well with human intuition because removing the set of edges identified by an explanation from the input graph changes the prediction significantly. Exhaustive experiments on many public datasets demonstrate the superior performance of our method.