Interpreting Graph Inference with Skyline Explanations

TL;DR

This paper presents skyline explanations, a multi-criteria interpretability framework for GNNs that generates diverse, comprehensive explanations by optimizing multiple explainability measures simultaneously.

Contribution

It introduces skyline explanations as a Pareto-based interpretability method, formulates it as a multi-criteria optimization, and develops scalable algorithms for large-scale graph analysis.

Findings

Effective in providing comprehensive GNN explanations

Scalable algorithms verified on real-world and synthetic graphs

Enriches interpretability with diverse explanations

Abstract

Inference queries have been routinely issued to graph machine learning models such as graph neural networks (GNNs) for various network analytical tasks. Nevertheless, GNN outputs are often hard to interpret comprehensively. Existing methods typically conform to individual pre-defined explainability measures (such as fidelity), which often leads to biased, ``one-side'' interpretations. This paper introduces skyline explanation, a new paradigm that interprets GNN outputs by simultaneously optimizing multiple explainability measures of users' interests. (1) We propose skyline explanations as a Pareto set of explanatory subgraphs that dominate others over multiple explanatory measures. We formulate skyline explanation as a multi-criteria optimization problem, and establish its hardness results. (2) We design efficient algorithms with an onion-peeling approach, which strategically prioritizes nodes and removes unpromising edges to incrementally assemble skyline explanations. (3) We also develop an algorithm to diversify the skyline explanations to enrich the comprehensive interpretation. (4) We introduce efficient parallel algorithms with load-balancing strategies to scale skyline explanation for large-scale GNN-based inference. Using real-world and synthetic graphs, we experimentally verify our algorithms' effectiveness and scalability.