|国家预印本平台
首页|Establishing validated standards for Home and Work location Detection

Establishing validated standards for Home and Work location Detection

Establishing validated standards for Home and Work location Detection

来源:Arxiv_logoArxiv
英文摘要

Smartphone location data have transformed urban mobility research, providing unprecedented insights into how people navigate and interact in cities. However, leveraging location data at scale presents methodological challenges. Accurately identifying individuals' home and work locations is critical for a range of applications, including commuting analysis, unemployment estimation, and urban accessibility studies. Despite their widespread use, home-work detection methods lack a standardized framework that accounts for differing data quality and that is validated against ground-truth observations. This limits the comparability and reproducibility of results across studies and datasets. In this paper, we present HoWDe, a robust algorithm for identifying home and work locations from mobility data, explicitly designed to handle missing data and varying data quality across individuals. Using two unique ground-truth datasets comprising over 5100 individuals from more than 80 countries, HoWDe achieves home and work detection accuracies of up to 97% and 88%, respectively, with consistent performance across countries and demographic groups. We examine how parameter choices shape the trade-off between accuracy and user retention, and demonstrate how these methodological decisions influence downstream applications such as employment estimation and commuting pattern analysis. By supporting in-house pre-processing through a transparent and validated pipeline, HoWDe also facilitates the sharing of privacy-preserving mobility data. Together, our tools and findings establish methodological standards that support more robust, scalable, and reproducible mobility research at both individual and urban scales.

Silvia de Sojo、Lorenzo Lucchini、Ollin D. Langle-Chimal、Samuel P. Fraiberger、Laura Alessandretti

计算技术、计算机技术

Silvia de Sojo,Lorenzo Lucchini,Ollin D. Langle-Chimal,Samuel P. Fraiberger,Laura Alessandretti.Establishing validated standards for Home and Work location Detection[EB/OL].(2025-06-23)[2025-07-17].https://arxiv.org/abs/2506.20679.点此复制

评论