Abstract
The rapid expansion of sensor technologies and location-enabled mobile applications has greatly advanced the study of human mobility. Recently emerged sources like mobile app data offer outputs similar to traditional datasets-trip chains, flows, and indicators-but the methods and decisions used to process such data often vary across location, time, and policy context and remain poorly documented and insufficiently transparent. This variability necessitates tailored data processing and validation approaches, which remain underexplored in existing literature. This study aims to provide a reproducible and replicable framework for processing location-points data, using a case study of anonymised mobility records collected in November 2021 across England. We describe a modular workflow and multi-stage validation techniques that enhance the reproducibility of stay-point detection and activity labelling. Furthermore, we demonstrate how the proposed framework can generate reliable mobility indicators and origin-destination flow matrices for broader research applications. The resulting datasets-including sampled anonymised trajectories and full origin-destination flow matrices-are publicly available for research purposes, with updates to code and methodology hosted on GitHub and Zenodo.