Human mobility prediction improved by spatially aware multi-agent llms
Enhancing Human Mobility Prediction with Spatially Aware LLM-based Multi-Agent Systems
Social and Information NetworksComputers and SocietyMultiagent Systems
Summary
Predicting the next place a person will visit is hard because it depends not just on past visits but also on distances and neighborhoods. The authors found that large language models (LLMs) often miss these spatial details, so they built a system where different agents focus on patterns, spatial reasoning, and making decisions together. By including real-world geography like distances and neighborhood info, their system predicts destinations better than older methods. This shows that understanding space explicitly helps computers guess where people will go next.
What this means in practice
- •For mobile app developers: Improve location-based recommendations by combining user behavior with real-world distance and neighborhood data for better prediction of next visited places.
- •For urban planners: Use improved next-location predictions that include spatial constraints to better understand and design urban mobility patterns.
Tested on one dataset.
Authors
Shangyu Lou, Ziqi Cui
Abstract
Predicting a user's next POI is a task in human mobility modeling, yet LLM-based approaches focus on semantic reasoning from previous mobility records, while neglecting real-world spatial context. However, human mobility is inherently shaped by spatial cognition, including geographic distance and neighborhood context. This issue is further compounded by prior evidence that LLMs often struggle with spatial reasoning tasks, including distance estimation and geographically biased prediction. To address these limitations, we propose our framework, a multi-agent LLM framework that decomposes next-POI prediction into three stages: Firstly, a Pattern Extraction Agent that captures temporal and categorical mobility patterns from trajectory history; Secondly, a Spatial Reasoning Agent that structures candidate activity choices by combining behavioral preferences with real-world spatial constraints, including geographic distance, road network distance, and neighborhood affiliation; and Thirdly, a Decision Synthesis Agent that integrates behavioral patterns and spatial reasoning for final prediction. Experiments on the NYC benchmark dataset with two LLM backbones show improvements over baseline methods, with up to 493% Hit@1 improvement and 37% relative improvement in Hit@5. Ablations show that combining neighborhood affiliation with distance-based features generally outperforms distance-only settings, and that the Spatial Reasoning Agent plays a crucial role in final prediction by integrating behavioral preferences with real-world spatial constraints, especially for smaller models. Overall, the results highlight the importance of spatial reasoning in mobility prediction. Accurate next-POI prediction requires combining behavioral patterns with explicit real-world spatial constraints, and multi-agent decomposition provides an effective structure for organizing these forms of context.