Hierarchical Attention-based Graph Neural Network with Relevance-driven Pruning

TL;DR

This paper introduces HA-HeteroGNN, a hierarchical attention-based GNN framework that enhances interpretability and efficiency by relevance-driven pruning, reducing graph complexity and improving classification accuracy.

Contribution

The paper presents a unified explainability-to-pruning pipeline for heterogeneous GNNs, utilizing a two-tier attention mechanism and relevance scores for effective node pruning.

Findings

27% reduction in graph edges through pruning

Improved classification accuracy by 2.4-6.1%

Up to 43.9% training-time reduction

Abstract

Graph Neural Networks (GNNs) excel at relational reasoning but face two persistent challenges: the lack of interpretable attribution for heterogeneous node types, and the computational overhead of message passing over large, noisy graphs. We propose the Hierarchical Attention-based Heterogeneous GNN (HA-HeteroGNN), a framework that addresses both issues through a unied explainability-to-pruning pipeline. A two-tier attention mechanism separates sensor-level and context-level computation across 16 node types and 18 edge types, producing per-node relevance scores via an attention-based GNN Explainer without requiring gradient backpropagation. These relevance scores then serve as a principled pruning criterion: removing nodes identied as consistently uninformative yields a 27% reduction in graph edges while simultaneously improving classication accuracy by 2.46.1% across all model variants, challenging the conventional assumption that pruning necessarily trades accuracy for eciency. Experiments on a 50,000-record synthetic dataset spanning 11 report categories demonstrate 97.5% cross-strategy explanation stability and domain consistent sensor attribution, with training-time reductions of up to 43.9% and real-time inference latency of approximately 5860 ms per sample.