CATS: Conditional Adversarial Trajectory Synthesis for privacy-preserving trajectory data publication using deep learning approaches

Abstract

AbstractThe prevalence of ubiquitous location-aware devices and mobile Internet enables us to collect massive individual-level trajectory dataset from users. Such trajectory big data bring new opportunities to human mobility research but also raise public concerns with regard to location privacy. In this work, we present the Conditional Adversarial Trajectory Synthesis (CATS), a deep-learning-based GeoAI methodological framework for privacy-preserving trajectory data generation and publication. CATS applies K-anonymity to the underlying spatiotemporal distributions of human movements, which provides a distributional-level strong privacy guarantee. By leveraging conditional adversarial training on K-anonymized human mobility matrices, trajectory global context learning using the attention-based mechanism, and recurrent bipartite graph matching of adjacent trajectory points, CATS is able to reconstruct trajectory topology from conditionally sampled locations and generate high-quality individual-level synthetic trajectory data, which can serve as supplements or alternatives to raw data for privacy-preserving trajectory data publication. The experiment results on over 90k GPS trajectories show that our method has a better performance in privacy preservation, spatiotemporal characteristic preservation, and downstream utility compared with baseline methods, which brings new insights into privacy-preserving human mobility research using generative AI techniques and explores data ethics issues in GIScience.Keywords: Geoprivacygenerative adversarial networkhuman mobilityGeoAIsynthetic data generation AcknowledgmentThe authors acknowledge the funding support provided by the American Family Insurance Data Science Institute Funding Initiative at the University of Wisconsin-Madison. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funder(s).Disclosure statementNo potential conflict of interest was reported by the author(s).Data and codes availability statementThe data and codes that support the findings of this study are available at the following link on figshare: https://doi.org/10.6084/m9.figshare.20760970. It is worth noting that due to the non-disclosure agreement with the data provider, we are not releasing the original individual-level GPS trajectory data but sharing the k-anonymized aggregated human mobility data used in our experiments.Additional informationNotes on contributorsJinmeng RaoJinmeng Rao is a research scientist at Mineral Earth Sciences. He received his PhD degree from the Department of Geography, University of Wisconsin-Madison. His research interests include GeoAI, Privacy-Preserving AI, and Location Privacy.Song GaoSong Gao is an associate professor in GIScience at the Department of Geography, University of Wisconsin-Madison. He holds a PhD in Geography at the University of California, Santa Barbara. His main research interests include place-based GIS, geospatial data science and GeoAI approaches to human mobility and social sensing.Sijia ZhuSijia Zhu is a Master student in Data Science at Columbia University. She received her bachelor degrees in Statistics and Economics from the University of Wisconsin-Madison.