Adaptive Location Hierarchy Learning for Long-Tailed Mobility Prediction

2025-05-26

AI Summary

To address poor tail-location prediction performance caused by long-tailed distribution in human mobility forecasting, this paper proposes ALOHA—a plug-and-play adaptive location hierarchy learning framework. Methodologically, ALOHA innovatively integrates Maslow’s motivation theory into chain-of-thought (CoT) prompting to guide large language models (LLMs) in capturing city-specific spatiotemporal semantics of locations; it further jointly optimizes head- and tail-location prediction within a tree-structured hierarchy via Gumbel-Softmax reparameterization and node-adaptive weighting. Evaluated on six real-world datasets, ALOHA significantly improves tail-location accuracy while preserving head-location performance. Ablation studies and weight analysis validate the synergistic effectiveness of its components, and case studies confirm the efficacy of semantics-driven hierarchical modeling.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Human mobility prediction is crucial for applications ranging from location-based recommendations to urban planning, which aims to forecast users' next location visits based on historical trajectories. Despite the severe long-tailed distribution of locations, the problem of long-tailed mobility prediction remains largely underexplored. Existing long-tailed learning methods primarily focus on rebalancing the skewed distribution at the data, model, or class level, neglecting to exploit the spatiotemporal semantics of locations. To address this gap, we propose the first plug-and-play framework for long-tailed mobility prediction in an exploitation and exploration manner, named extbf{A}daptive extbf{LO}cation extbf{H}ier extbf{A}rchy learning (ALOHA). First, we construct city-tailored location hierarchy based on Large Language Models (LLMs) by exploiting Maslow's theory of human motivation to design Chain-of-Thought (CoT) prompts that captures spatiotemporal semantics. Second, we optimize the location hierarchy predictions by Gumbel disturbance and node-wise adaptive weights within the hierarchical tree structure. Experiments on state-of-the-art models across six datasets demonstrate the framework's consistent effectiveness and generalizability, which strikes a well balance between head and tail locations. Weight analysis and ablation studies reveal the optimization differences of each component for head and tail locations. Furthermore, in-depth analyses of hierarchical distance and case study demonstrate the effective semantic guidance from the location hierarchy. Our code will be made publicly available.

Problem

• Addresses long-tailed mobility prediction neglect in existing methods
• Exploits spatiotemporal semantics using adaptive location hierarchy
• Balances prediction accuracy for head and tail locations

Innovation

• LLM-based location hierarchy construction
• Gumbel disturbance for hierarchy optimization
• Node-wise adaptive weights in tree structure