Message-passing selection: Towards interpretable GNNs for graph classification

TL;DR

This paper introduces MSInterpreter, an interpretable message-passing scheme for GNNs that selects critical message paths for self-explanation, enhancing interpretability in graph classification tasks.

Contribution

The paper proposes MSScheme, a novel message-passing selection method that provides self-explainability for GNNs, applicable as a plug-and-play module across various models.

Findings

Effective on graph classification benchmarks

Improves interpretability of GNNs

Demonstrates self-explainability in message passing

Abstract

In this paper, we strive to develop an interpretable GNNs' inference paradigm, termed MSInterpreter, which can serve as a plug-and-play scheme readily applicable to various GNNs' baselines. Unlike the most existing explanation methods, MSInterpreter provides a Message-passing Selection scheme(MSScheme) to select the critical paths for GNNs' message aggregations, which aims at reaching the self-explaination instead of post-hoc explanations. In detail, the elaborate MSScheme is designed to calculate weight factors of message aggregation paths by considering the vanilla structure and node embedding components, where the structure base aims at weight factors among node-induced substructures; on the other hand, the node embedding base focuses on weight factors via node embeddings obtained by one-layer GNN.Finally, we demonstrate the effectiveness of our approach on graph classification benchmarks.