MoveFM-R: Advancing Mobility Foundation Models via Language-driven Semantic Reasoning

2025-09-26

AI Summary

Existing mobile foundation models (MFMs) suffer from limited training data scale and insufficient semantic modeling capability, while large language models (LLMs), though powerful in semantic reasoning, lack spatiotemporal and physical constraints, hindering realistic trajectory generation. To address this, we propose a language-driven mobile modeling framework that achieves the first deep integration of LLMs’ semantic understanding with MFMs’ spatiotemporal statistical modeling. Specifically, we design a semantic-enhanced positional encoding to bridge the representational gap between geographic coordinates and linguistic tokens; introduce a progressive alignment curriculum and an interactive self-reflection mechanism to improve cross-modal robustness; and formulate a conditional trajectory generation paradigm enabling end-to-end mapping from natural language instructions to realistic trajectories. Extensive evaluations across multiple benchmarks demonstrate significant improvements over state-of-the-art methods, with strong zero-shot generalization capability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Planning, Routing, and Scheduling: Planning with Language Models

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: Large language models for search

Abstract

Mobility Foundation Models (MFMs) have advanced the modeling of human movement patterns, yet they face a ceiling due to limitations in data scale and semantic understanding. While Large Language Models (LLMs) offer powerful semantic reasoning, they lack the innate understanding of spatio-temporal statistics required for generating physically plausible mobility trajectories. To address these gaps, we propose MoveFM-R, a novel framework that unlocks the full potential of mobility foundation models by leveraging language-driven semantic reasoning capabilities. It tackles two key challenges: the vocabulary mismatch between continuous geographic coordinates and discrete language tokens, and the representation gap between the latent vectors of MFMs and the semantic world of LLMs. MoveFM-R is built on three core innovations: a semantically enhanced location encoding to bridge the geography-language gap, a progressive curriculum to align the LLM's reasoning with mobility patterns, and an interactive self-reflection mechanism for conditional trajectory generation. Extensive experiments demonstrate that MoveFM-R significantly outperforms existing MFM-based and LLM-based baselines. It also shows robust generalization in zero-shot settings and excels at generating realistic trajectories from natural language instructions. By synthesizing the statistical power of MFMs with the deep semantic understanding of LLMs, MoveFM-R pioneers a new paradigm that enables a more comprehensive, interpretable, and powerful modeling of human mobility. The implementation of MoveFM-R is available online at https://anonymous.4open.science/r/MoveFM-R-CDE7/.

Problem

• Bridging vocabulary mismatch between geographic coordinates and language tokens
• Addressing representation gap between mobility models and semantic understanding
• Enhancing trajectory generation through language-driven semantic reasoning capabilities

Innovation

• Semantically enhanced location encoding bridges geography-language gap
• Progressive curriculum aligns LLM reasoning with mobility patterns
• Interactive self-reflection mechanism enables conditional trajectory generation