DEGREE: Decomposition Based Explanation For Graph Neural Networks

TL;DR

DEGREE offers a faithful, decomposition-based explanation method for GNNs, enabling detailed understanding of input contributions and complex node interactions, thereby improving interpretability for graph classification tasks.

Contribution

This paper introduces DEGREE, a novel decomposition approach that enhances GNN interpretability by tracking component contributions and revealing complex node interactions.

Findings

DEGREE provides faithful explanations for GNN predictions.

The method effectively uncovers complex node interactions.

Experiments show improved interpretability on real-world datasets.

Abstract

Graph Neural Networks (GNNs) are gaining extensive attention for their application in graph data. However, the black-box nature of GNNs prevents users from understanding and trusting the models, thus hampering their applicability. Whereas explaining GNNs remains a challenge, most existing methods fall into approximation based and perturbation based approaches with suffer from faithfulness problems and unnatural artifacts, respectively. To tackle these problems, we propose DEGREE \degree to provide a faithful explanation for GNN predictions. By decomposing the information generation and aggregation mechanism of GNNs, DEGREE allows tracking the contributions of specific components of the input graph to the final prediction. Based on this, we further design a subgraph level interpretation algorithm to reveal complex interactions between graph nodes that are overlooked by previous methods. The efficiency of our algorithm can be further improved by utilizing GNN characteristics. Finally, we conduct quantitative and qualitative experiments on synthetic and real-world datasets to demonstrate the effectiveness of DEGREE on node classification and graph classification tasks.