Explainable Hierarchical Graph Neural Networks for Structured Decision Modeling

Abstract

Decision modeling in complex systems involves heterogeneous factors with different degrees of controllability and structural dependence, yet most neural models treat all inputs uniformly and provide limited decision-level interpretability. We propose an Explainable Hierarchical Graph Neural Network (EH-GNN) that decomposes inputs into contextual, controllable, and structural components. Contextual and controllable variables are modeled using hierarchical neural encoders, while structural dependencies among categorical entities are captured using a graph neural network. The model supports component-level attribution, enabling explanations to be aligned with actionable and non-actionable decision factors rather than individual features. Unlike post-hoc explanation methods, EH-GNN integrates this component-level attribution
directly into the model architecture, enabling stable and decision-aligned explanations by design.
We evaluate the proposed framework on the publicly available Rossmann sales benchmark dataset, which exhibits strong structural heterogeneity and relational effects. EH-GNN is compared against non-hierarchical multilayer perceptrons, graph-only neural models, and
post-hoc explainability pipelines based on feature attribution methods. Experimental results show that the proposed approach achieves competitive predictive performance while producing stable and semantically coherent attributions that are not recoverable from conventional
baselines. While not optimized solely for error minimization, the proposed framework emphasizes decision-aligned interpretability alongside predictive accuracy.