Extracting Interpretable Logic Rules from Graph Neural Networks

2025-03-25

AI Summary

Graph Neural Networks (GNNs) suffer from poor interpretability—operating as “black boxes”—hindering their adoption in high-stakes domains like drug discovery. Method: We propose LOGICXGNN, the first model-agnostic, data-driven framework for automatic extraction of first-order logical rules from GNNs. It integrates symbolic inductive logic programming, differentiable logical reasoning, and GNN feature distillation—requiring no predefined domain knowledge—to derive global, concept-level explanations. Contribution/Results: The extracted rules achieve superior discriminative performance (outperforming base GNNs on MUTAG, BBBP, and other benchmarks) and strong generative capability, enabling counterfactual graph synthesis and controllable molecular structure design. Experiments uncover previously overlooked chemical substructure patterns—ignored by mainstream methods—and demonstrate cross-task rule transferability. LOGICXGNN establishes a new paradigm for trustworthy, AI-driven drug discovery.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Logic Programming
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph neural networks (GNNs) operate over both input feature spaces and combinatorial graph structures, making it challenging to understand the rationale behind their predictions. As GNNs gain widespread popularity and demonstrate success across various domains, such as drug discovery, studying their interpretability has become a critical task. To address this, many explainability methods have been proposed, with recent efforts shifting from instance-specific explanations to global concept-based explainability. However, these approaches face several limitations, such as relying on predefined concepts and explaining only a limited set of patterns. To address this, we propose a novel framework, LOGICXGNN, for extracting interpretable logic rules from GNNs. LOGICXGNN is model-agnostic, efficient, and data-driven, eliminating the need for predefined concepts. More importantly, it can serve as a rule-based classifier and even outperform the original neural models. Its interpretability facilitates knowledge discovery, as demonstrated by its ability to extract detailed and accurate chemistry knowledge that is often overlooked by existing methods. Another key advantage of LOGICXGNN is its ability to generate new graph instances in a controlled and transparent manner, offering significant potential for applications such as drug design. We empirically demonstrate these merits through experiments on real-world datasets such as MUTAG and BBBP.

Problem

• Extracting interpretable logic rules from GNNs
• Overcoming limitations of predefined concept-based explainability
• Generating controlled graph instances for applications like drug design

Innovation

• Extracts interpretable logic rules from GNNs
• Model-agnostic and data-driven framework
• Generates new graph instances transparently