Abstract
Estimating origin-destination (OD) flows under disruptive events is crucial for disaster response and urban resilience. Existing deep OD models trained on routine mobility often degrade when extreme events abruptly alter regional functions and population activities, while retraining a new generator for each event is impractical under limited event-time supervision. We propose EventOD, an event-adaptive OD generation framework that steers a pretrained OD generator using structured event semantics.
EventOD first uses a large language model to infer region-level functional and demographic control vectors from coarse event observations. It then learns two lightweight adaptation modules, AlphaNet and BetaNet, to calibrate the magnitude of these semantic shifts and further introduces a retrieval-augmented fallback pathway for scenarios with sparse supervision. The resulting event-conditioned features are injected into a pretrained graph diffusion OD model through input-level modulation, enabling event-aware adaptation without updating generator parameters.
Experiments on hurricane- and pandemic-induced mobility across U.S. counties show that EventOD consistently improves both reconstruction accuracy and distributional fidelity over strong baselines. Source code is available at EventOD GitHub.
Blogger's Review: The innovation of EventOD lies in integrating large language models with OD flow generation, addressing the adaptability issues of traditional models under extreme events. Through lightweight modules and augmented retrieval, EventOD effectively enhances model accuracy and robustness, extending deep learning applications to disaster response and urban planning. Its methodology offers a new perspective for future event-driven models, warranting further exploration and practice.