Encoding Concepts in Graph Neural Networks

TL;DR

This paper introduces a differentiable concept-discovery module for graph neural networks that enhances interpretability, trust, and performance without sacrificing accuracy.

Contribution

It presents the first differentiable concept encoder for graph networks, enabling explainability by design through concept discovery and utilization.

Findings

Achieves comparable accuracy to vanilla graph networks.

Discovers meaningful, high-quality concepts with high completeness and purity.

Provides effective, high-quality concept-based explanations and interventions.

Abstract

The opaque reasoning of Graph Neural Networks induces a lack of human trust. Existing graph network explainers attempt to address this issue by providing post-hoc explanations, however, they fail to make the model itself more interpretable. To fill this gap, we introduce the Concept Encoder Module, the first differentiable concept-discovery approach for graph networks. The proposed approach makes graph networks explainable by design by first discovering graph concepts and then using these to solve the task. Our results demonstrate that this approach allows graph networks to: (i) attain model accuracy comparable with their equivalent vanilla versions, (ii) discover meaningful concepts that achieve high concept completeness and purity scores, (iii) provide high-quality concept-based logic explanations for their prediction, and (iv) support effective interventions at test time: these can increase human trust as well as significantly improve model performance.