Activity-aware Human Mobility Prediction with Hierarchical Graph Attention Recurrent Network
Yihong Tang, Junlin He, Zhan Zhao
TL;DR
This work tackles next location prediction by explicitly incorporating activity information and complex time-activity-location dependencies. It introduces HGARN, a Hierarchical Graph Attention Recurrent Network that builds a three-layer activity-aware graph (location, localized-activity, activity) and uses a Temporal Module to jointly predict next activities and locations, aided by a model-agnostic MaHec label that leverages user history. Key contributions include the hierarchical graph construction with location-activity couplings, multi-layer graph attention learning, the MaHec soft-label mechanism, and a fusion-based temporal decoder. Empirical results on two real-world LBSN datasets show state-of-the-art performance in both recurring and explorative settings, with interpretable attention patterns that reveal plausible activity dependencies. The approach offers a practical, scalable framework for activity-aware transportation analytics and personalized location recommendations, with code and resources made available for reproducibility.
Abstract
Human mobility prediction is a fundamental task essential for various applications in urban planning, location-based services and intelligent transportation systems. Existing methods often ignore activity information crucial for reasoning human preferences and routines, or adopt a simplified representation of the dependencies between time, activities and locations. To address these issues, we present Hierarchical Graph Attention Recurrent Network (HGARN) for human mobility prediction. Specifically, we construct a hierarchical graph based on past mobility records and employ a Hierarchical Graph Attention Module to capture complex time-activity-location dependencies. This way, HGARN can learn representations with rich human travel semantics to model user preferences at the global level. We also propose a model-agnostic history-enhanced confidence (MAHEC) label to incorporate each user's individual-level preferences. Finally, we introduce a Temporal Module, which employs recurrent structures to jointly predict users' next activities and their associated locations, with the former used as an auxiliary task to enhance the latter prediction. For model evaluation, we test the performance of HGARN against existing state-of-the-art methods in both the recurring (i.e., returning to a previously visited location) and explorative (i.e., visiting a new location) settings. Overall, HGARN outperforms other baselines significantly in all settings based on two real-world human mobility data benchmarks. These findings confirm the important role that human activities play in determining mobility decisions, illustrating the need to develop activity-aware intelligent transportation systems. Source codes of this study are available at https://github.com/YihongT/HGARN.