ELLMob: Event-Driven Human Mobility Generation with Self-Aligned LLM Framework

TL;DR

ELLMob introduces a novel self-aligned LLM framework for human mobility generation that effectively captures both routine and event-driven trajectories, supported by a new event-annotated dataset and superior experimental results.

Contribution

The paper presents the first event-annotated mobility dataset and a self-aligned LLM framework that balances habitual patterns and event constraints for realistic trajectory synthesis.

Findings

ELLMob outperforms state-of-the-art methods across all tested events.

The constructed dataset covers Typhoon Hagibis, COVID-19, and Tokyo 2021 Olympics.

The framework effectively models both routine and deviated human mobility behaviors.

Abstract

Human mobility generation aims to synthesize plausible trajectory data, which is widely used in urban system research. While Large Language Model-based methods excel at generating routine trajectories, they struggle to capture deviated mobility during large-scale societal events. This limitation stems from two critical gaps: (1) the absence of event-annotated mobility datasets for design and evaluation, and (2) the inability of current frameworks to reconcile competitions between users' habitual patterns and event-imposed constraints when making trajectory decisions. This work addresses these gaps with a twofold contribution. First, we construct the first event-annotated mobility dataset covering three major events: Typhoon Hagibis, COVID-19, and the Tokyo 2021 Olympics. Second, we propose ELLMob, a self-aligned LLM framework that first extracts competing rationales between habitual patterns and event constraints, based on Fuzzy-Trace Theory, and then iteratively aligns them to generate trajectories that are both habitually grounded and event-responsive. Extensive experiments show that ELLMob wins state-of-the-art baselines across all events, demonstrating its effectiveness. Our codes and datasets are available at https://github.com/deepkashiwa20/ELLMob.