Identifying home locations in human mobility data: an open-source R package for comparison and reproducibility

Abstract

Identifying meaningful locations, such as home or work, from human mobility data has become an increasingly common prerequisite for geographic research. Although location-based services (LBS) and other mobile technology have rapidly grown in recent years, it can be challenging to infer meaningful places from such data, whichcompared to conventional datasetscan be devoid of context. Existing approaches are often developed ad-hoc and can lack transparency and reproducibility. To address this, we introduce an R software package for inferring home locations from LBS data. The package implements pre-existing algorithms and provides building blocks to make writing algorithmic 'recipes' more convenient. We evaluate this approach by analyzing a de-identified LBS dataset from Singapore that aims to balance ethics and privacy with the research goal of identifying meaningful locations. We show that ensemble approaches, combining multiple algorithms, can be especially valuable in this regard as the resulting patterns of inferred home locations closely correlate with the distribution of residential population. We hope this package, and others like it, will contribute to an increase in use and sharing of comparable algorithms, research code and data. This will increase transparency

PrivyTo: A privacy‐preserving location‐sharing platform

“…A number of tools have been developed with the goal of allowing users to mask location data. Swanlund, Schuurman and Brussoni (2020) constructed MaskMy.XYZ , a tool for obfuscating geographic data sets through an online web application, and Chen and Poorthuis (2021) developed an R package for identifying (and obfuscating) home locations from mobility data. Drakonakis, Ilia, Ioannidis, &, Polakis (2019) find that if users are given the choice to explicitly select what location data they publish on social networking applications, there is a 95% reduction in attaching their coordinates to their posts.…”

Monitoring streets through tweets: Using user-generated geographic information to predict gentrification and displacement

“…In this sense, we have designed this procedure to prevent false positives (cf. Chen and Poorthuis, 2021 for a sensitivity analysis and discussion of different home location algorithms).…”

Entangled footprints: Understanding urban neighbourhoods by measuring distance, diversity, and direction of flows in Singapore

“…Considering the varying shapes and sizes of administrative regions, and balancing privacy and ethics with keeping sufficient observations in individual neighbourhoods for subsequent analyses, we have chosen to use a grid cell with 750m resolution (cf. Chen and Poorthuis (2021) for a discussion of this scale) as the unit of analysis.…”

“…The de-identified LBS dataset used for this analysis consists of around 20.9 million tweets created by approximately 130.3k users sent from Singapore between 2012 and 2016 (Chen and Poorthuis, 2021). Each data point includes three attributes -a unique identifier for the user; a unique identifier for the spatial neighbourhood for the data point; and a timestamp that reflects the time the data point was created.…”

“…For our analysis, we utilize a de-identified dataset that consists of all tweets sent between 2012 and 2016 from Singapore (Chen and Poorthuis, 2021). Although we use Twitter data in our case, the methodology proposed here can be applied to data from other Location-Based Services (LBS), such as Instagram, cell phone data, Weibo, Facebook, etc.…”