Where to Go Next Day: Multi-scale Spatial-Temporal Decoupled Model for Mid-term Human Mobility Prediction

2025-01-11

AI Summary

Existing models predominantly focus on short-term (next-hop) mobility prediction, failing to support applications requiring medium- to long-term (1-day–1-week) individual trajectory modeling—such as traffic dispatching and epidemic simulation. This paper proposes the first dedicated framework for medium- to long-term human mobility forecasting. It introduces a spatiotemporal decoupling architecture that decomposes daily trajectories into location sequences and dwell-duration sequences. A multi-scale spatiotemporal hierarchical encoder and a spatially heterogeneous graph learner jointly capture daily recurrence, weekly periodicity, and cross-regional semantic correlations. The model integrates a Transformer-based decoder with hierarchical temporal encoding, synergizing graph neural networks and statistical physics–inspired analysis. Evaluated on mobile signaling data from five cities, it significantly outperforms state-of-the-art methods. Applied to COVID-19 spread simulation in Boston, it reduces the MAE of cumulative new case predictions by 62.8%.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Predicting individual mobility patterns is crucial across various applications. While current methods mainly focus on predicting the next location for personalized services like recommendations, they often fall short in supporting broader applications such as traffic management and epidemic control, which require longer period forecasts of human mobility. This study addresses mid-term mobility prediction, aiming to capture daily travel patterns and forecast trajectories for the upcoming day or week. We propose a novel Multi-scale Spatial-Temporal Decoupled Predictor (MSTDP) designed to efficiently extract spatial and temporal information by decoupling daily trajectories into distinct location-duration chains. Our approach employs a hierarchical encoder to model multi-scale temporal patterns, including daily recurrence and weekly periodicity, and utilizes a transformer-based decoder to globally attend to predicted information in the location or duration chain. Additionally, we introduce a spatial heterogeneous graph learner to capture multi-scale spatial relationships, enhancing semantic-rich representations. Extensive experiments, including statistical physics analysis, are conducted on large-scale mobile phone records in five cities (Boston, Los Angeles, SF Bay Area, Shanghai, and Tokyo), to demonstrate MSTDP's advantages. Applied to epidemic modeling in Boston, MSTDP significantly outperforms the best-performing baseline, achieving a remarkable 62.8% reduction in MAE for cumulative new cases.

Problem

• long-term mobility prediction
• traffic management
• epidemic control

Innovation

• MSTDP
• Graph Learner
• Epidemic Prediction