Generating Skyline Explanations for Graph Neural Networks

2025-05-12

AI Summary

Existing GNN explanation methods typically optimize a single objective (e.g., fidelity or sparsity), yielding partial, biased explanations. Method: This paper introduces the “skyline explanation” paradigm—the first to formulate GNN interpretability as a multi-objective optimization problem, jointly optimizing fidelity, sparsity, diversity, and other criteria to generate a Pareto-optimal set of high-quality subgraph explanations. We formally prove the task is NP-hard and propose the theoretically grounded “onion-peeling” edge-removal algorithm, augmented with heuristic pruning and diversity-enhancing mechanisms. Contribution/Results: Extensive evaluation on real-world graph datasets demonstrates that our method achieves superior efficiency, scalability, and comprehensive explanation quality. It significantly improves user trust in GNN decisions by providing balanced, non-dominated explanations that capture multiple complementary aspects of model behavior.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Search and Optimization: Combinatorial Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

This paper proposes a novel approach to generate subgraph explanations for graph neural networks GNNs that simultaneously optimize multiple measures for explainability. Existing GNN explanation methods often compute subgraphs (called ``explanatory subgraphs'') that optimize a pre-defined, single explainability measure, such as fidelity or conciseness. This can lead to biased explanations that cannot provide a comprehensive explanation to clarify the output of GNN models. We introduce skyline explanation, a GNN explanation paradigm that aims to identify k explanatory subgraphs by simultaneously optimizing multiple explainability measures. (1) We formulate skyline explanation generation as a multi-objective optimization problem, and pursue explanations that approximate a skyline set of explanatory subgraphs. We show the hardness for skyline explanation generation. (2) We design efficient algorithms with an onion-peeling approach that strategically removes edges from neighbors of nodes of interests, and incrementally improves explanations as it explores an interpretation domain, with provable quality guarantees. (3) We further develop an algorithm to diversify explanations to provide more comprehensive perspectives. Using real-world graphs, we empirically verify the effectiveness, efficiency, and scalability of our algorithms.

Problem

• Generates subgraph explanations optimizing multiple explainability measures
• Addresses bias in single-measure GNN explanation methods
• Proposes efficient algorithms with quality guarantees for skyline explanations

Innovation

• Multi-objective optimization for GNN explanations
• Onion-peeling algorithm with quality guarantees
• Diversified subgraph explanations for comprehensiveness