PHGNet: Prototype-Guided Hypergraph Construction for Heterogeneous Spatiotemporal Forecasting

2026-05-25

AI Summary

This work addresses the challenge of effectively modeling high-order dynamic spatiotemporal dependencies with spatial heterogeneity in traffic forecasting. To this end, the authors propose PHGNet, a novel framework that introduces prototype learning to guide the construction of dynamic hypergraphs, adaptively aggregating nodes with similar traffic patterns into time-varying hyperedges. The framework further incorporates a global–local node representation module to enhance hypergraph reliability and combines iterative residual refinement with a temporal query attention mechanism to enable efficient parallel decoding. Extensive experiments on multiple real-world traffic datasets demonstrate that PHGNet significantly outperforms state-of-the-art methods, achieving notably higher prediction accuracy.

Abstract

As a core task in intelligent transportation systems, traffic forecasting plays a critical role in urban traffic management. Accurate traffic forecasting relies on modeling complex spatiotemporal dependencies, which is inherently challenging due to spatial heterogeneity in traffic systems.Despite significant progress, most existing methods are still limited to pairwise spatial dependency modeling, making it difficult to capture dynamic high-order interactions among nodes with similar traffic patterns. To address this issue, we propose PHGNet, a novel spatiotemporal forecasting framework based on prototype-guided hypergraph construction. At the core of PHGNet, a prototype learning mechanism is designed to adaptively assign pattern-similar nodes to hyperedges, thereby capturing high-order interactions with time-varying structures. To improve the reliability of dynamic hypergraph construction, we further develop a global-local node representation module to extract time-consistent features. For forecasting, iterative residual refinement and Temporal Query Attention are introduced to improve forecasting accuracy while supporting efficient parallel decoding. Extensive experiments on multiple real-world datasets demonstrate that PHGNet achieves superior predictive performance compared with state-of-the-art methods.

Problem

• spatiotemporal forecasting
• spatial heterogeneity
• high-order interactions
• traffic prediction
• dynamic hypergraph

Innovation

• hypergraph construction
• prototype learning
• spatiotemporal forecasting
• high-order interactions
• heterogeneous traffic systems