GnnXemplar: Exemplars to Explanations -- Natural Language Rules for Global GNN Interpretability
This addresses the need for trustworthy GNNs in real-world applications by providing global explanations, though it is incremental as it builds on existing global explanation methods with a novel approach.
The paper tackled the problem of global interpretability for Graph Neural Networks (GNNs) in node classification, proposing GnnXemplar, which uses exemplars and natural language rules to explain predictions, resulting in significant outperformance over existing methods in fidelity, scalability, and human interpretability as validated by a user study with 60 participants.
Graph Neural Networks (GNNs) are widely used for node classification, yet their opaque decision-making limits trust and adoption. While local explanations offer insights into individual predictions, global explanation methods, those that characterize an entire class, remain underdeveloped. Existing global explainers rely on motif discovery in small graphs, an approach that breaks down in large, real-world settings where subgraph repetition is rare, node attributes are high-dimensional, and predictions arise from complex structure-attribute interactions. We propose GnnXemplar, a novel global explainer inspired from Exemplar Theory from cognitive science. GnnXemplar identifies representative nodes in the GNN embedding space, exemplars, and explains predictions using natural language rules derived from their neighborhoods. Exemplar selection is framed as a coverage maximization problem over reverse k-nearest neighbors, for which we provide an efficient greedy approximation. To derive interpretable rules, we employ a self-refining prompt strategy using large language models (LLMs). Experiments across diverse benchmarks show that GnnXemplar significantly outperforms existing methods in fidelity, scalability, and human interpretability, as validated by a user study with 60 participants.