Chen ChuHengcai ZhangPeixiao WangFeng Lua State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing,Chinab University of Chinese Academy of Sciences,Beijing,ChinaChen Chu is a Master candidate at State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences. His research interests focus on interpretable spatiotemporal prediction,human mobility,and spatiotemporal data mining. He contributed to the conception of the research idea,algorithm implementation,visualization,and the writing of the manuscript. cai Zhang is an Associate Professor at the Institute of Geographical Sciences and Natural Resources Research,Chinese Academy of Sciences. He received his Ph.D. degree from the Institute of Geographical Sciences and Natural Resources Research,Chinese Academy of Sciences in 2013. His research interest mainly focuses on 3D computing,urban computing and spatial-temporal data mining. He contributed to the supervision of the experiments,manuscript revision,and the writing of the manuscript. zhanghc@lreis.ac.cn.Peixiao Wang is a Postdoctoral Fellow from State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Nature Resources Research,Chinese Academy of Sciences. He received Ph.D. degree under from State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,and received the M.S. degree from The Academy of Digital China,Fuzhou University. His research topics include spatiotemporal data mining,and spatiotemporal prediction,especially focus on spatiotemporal prediction of transportation systems. He provided the computational resource and contributed to the manuscript revision. wpx@lreis.ac.cn.Feng Lu is a Professor at the Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences. His research interests cover trajectory data mining,computational transportation science and location-based services. He contributed to the conception of the research idea,manuscript revision and supervision. luf@lreis.ac.cn.

Abstract:Most mobility modeling methods are designed to solve specific tasks, leading to questions regarding their deficiency in generalizability. Inspired by the bloom of foundation models, we proposed a Trajectory Generation framework based on the Diffusion Model (TrajGDM) to capture the universal mobility pattern in a trajectory dataset by learning the trajectory generation process. The process is modeled as a step-by-step uncertainty-reducing process, in which a deep learning network with a novel training method is proposed to learn from the process. We compared the proposed trajectory generation method with six baselines on two public trajectory datasets. The results showed that the similarity between the generated and real trajectory movements measured by the Jensen-Shannon Divergence improved significantly on both datasets. Moreover, we applied zero-shot inferences on two basic trajectory tasks: trajectory prediction and trajectory reconstruction. The accuracy improved by a maximum of 25.6% on two tasks. The universal mobility pattern that is suitable for solving multiple trajectory tasks is verified, inferring the strong generalizability of our model. Finally, the study provides insights into artificial intelligence's understanding of human mobility by exploring the way the model maps the trajectory in the latent space into reality.

Regions are Who Walk Them: a Large Pre-trained Spatiotemporal Model Based on Human Mobility for Ubiquitous Urban Sensing

A Multi-Scale Unified Model of Human Mobility in Urban Agglomerations

MobiSeg: Interactive region segmentation using heterogeneous mobility data

UrbanMotion: Visual Analysis of Metropolitan-Scale Sparse Trajectories

Exploring Multi-Scale Spatiotemporal Twitter User Mobility Patterns with A Visual-Analytics Approach

Large-Scale Human Mobility Prediction Based on Periodic Attenuation and Local Feature Match

Human Mobility Prediction Using Sparse Trajectory Data.

Urban Region Representation Learning with Human Trajectories: a Multi-View Approach Incorporating Transition, Spatial, and Temporal Perspectives

Crowdsourced mobility prediction based on spatio-temporal contexts

Region Partition Using User Mobility Patterns Based on Topic Model

Deep ROI-Based Modeling for Urban Human Mobility Prediction

A Spatial-Temporal Topic Segmentation Model For Human Mobile Behavior

Simulating human mobility with a trajectory generation framework based on diffusion model

Semantics-Aware Hidden Markov Model for Human Mobility

Exploring trajectory embedding via spatial-temporal propagation for dynamic region representations

Pretrained Mobility Transformer: A Foundation Model for Human Mobility

A System of Mining Semantic Trajectory Patterns from GPS Data of Real Users

Modeling urban scale human mobility through big data analysis and machine learning

Deciphering Human Mobility: Inferring Semantics of Trajectories with Large Language Models

A Regionalization Method for Clustering and Partitioning Based on Trajectories from NLP Perspective

Discovering Social Relationship Between City Regions Using Human Mobility