Enhancing Large Language Models for Mobility Analytics with Semantic Location Tokenization

2025-06-08

AI Summary

Current large language models (LLMs) face two key bottlenecks in urban mobility data analysis: (1) locations are represented solely as discrete IDs, lacking semantic richness; and (2) mobility sequence modeling relies on single-objective instruction tuning, limiting representational capacity. To address these, we propose QT-Mob—a novel framework featuring a semantics-aware location tokenization module that maps raw geographic coordinates to compact, context-sensitive, learnable tokens; and a multi-objective collaborative fine-tuning mechanism that jointly optimizes sequential prediction and geospatial embedding learning, enabling internal representation alignment within the LLM. Evaluated on three real-world trajectory datasets, QT-Mob achieves significant improvements in next-location prediction and trajectory reconstruction, consistently outperforming state-of-the-art deep learning and LLM-based baselines. Our approach establishes a new paradigm for mobility intelligence—delivering both high accuracy and model interpretability through semantically grounded, large-scale language modeling.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

The widespread adoption of location-based services has led to the generation of vast amounts of mobility data, providing significant opportunities to model user movement dynamics within urban environments. Recent advancements have focused on adapting Large Language Models (LLMs) for mobility analytics. However, existing methods face two primary limitations: inadequate semantic representation of locations (i.e., discrete IDs) and insufficient modeling of mobility signals within LLMs (i.e., single templated instruction fine-tuning). To address these issues, we propose QT-Mob, a novel framework that significantly enhances LLMs for mobility analytics. QT-Mob introduces a location tokenization module that learns compact, semantically rich tokens to represent locations, preserving contextual information while ensuring compatibility with LLMs. Furthermore, QT-Mob incorporates a series of complementary fine-tuning objectives that align the learned tokens with the internal representations in LLMs, improving the model's comprehension of sequential movement patterns and location semantics. The proposed QT-Mob framework not only enhances LLMs' ability to interpret mobility data but also provides a more generalizable approach for various mobility analytics tasks. Experiments on three real-world dataset demonstrate the superior performance in both next-location prediction and mobility recovery tasks, outperforming existing deep learning and LLM-based methods.

Problem

• Inadequate semantic representation of locations in mobility data
• Insufficient modeling of mobility signals within LLMs
• Need for better LLM adaptation for mobility analytics tasks

Innovation

• Semantic location tokenization for LLMs
• Complementary fine-tuning objectives for mobility
• Enhanced LLM comprehension of movement patterns