A geographic data science framework for the functional and contextual analysis of human dynamics within global cities
Alessia Calafiore; Gregory Palmer; Sam Comber; Daniel Arribas-Bel; Alex Singleton (2021). Computers, Environment and Urban Systems, 85, 101539. DOI: 10.1016/j.compenvurbsys.2020.101539
Abstract
This study develops a Geographic Data Science framework that transforms the Foursquare check-in locations and user origin-destination flows data into knowledge about the emerging forms and characteristics of cities’ neighbourhoods. We employ a longitudinal mobility dataset describing human interactions with Foursquare venues in ten global cities: Chicago, Istanbul, Jakarta, London, Los Angeles, New York, Paris, Seoul, Singapore, Tokyo. This social media data provides spatio-temporally referenced digital traces left by human use of urban environments, giving us access to the intangible aspects of urban life, such as people behaviours and preferences. Our framework capitalizes on these new data sources, bringing about a novel Geographic Data Science and human-centered methodological approach. Combining network science – a study area with great promise for the analysis of cities and their structure – with geospatial analysis methods, we model cities as a series of global urban networks. Through a spatially weighted community detection algorithm, we uncover functional neighbourhoods for the ten global cities. Each neighbourhood is linked to hyper-local characterisations of their built environment for the Foursquare venues that compose them, and complemented with a range of measures describing their diversity, morphology and mobility. This information is used in a clustering exercise that uncovers a set of four functional neighbourhood types. Our results enable the profiling and comparison of functional neighbourhoods, based on human dynamics and their contexts, across the sample of global cities. The framework is portable to other geographic contexts where interaction data are available to bind different localities into functional agglomerations, and provide insight into their contextual and human dynamics.
Extended Summary
This research develops a novel geographic data science framework to identify and characterise functional neighbourhoods in global cities using social media location data. The study addresses how human mobility patterns and venue interactions can reveal the underlying structure and characteristics of urban neighbourhoods across diverse global contexts. The methodology combines network science techniques with geospatial analysis methods, utilising longitudinal Foursquare check-in data from ten major cities: Chicago, Istanbul, Jakarta, London, Los Angeles, New York, Paris, Seoul, Singapore, and Tokyo. The research employs a spatially weighted community detection algorithm to identify functional neighbourhoods based on user movement patterns between venues, creating geographic units that emerge from human behaviour rather than administrative boundaries. Each neighbourhood is then characterised using three key dimensions: diversity (measured through venue category variety using Shannon Index), context (including built environment features and temporal check-in patterns), and mobility (analysed through distance decay functions). The framework generates catchment areas around each venue using ten-minute walking distances and incorporates street network topology, density, and connectivity measures from OpenStreetMap data to capture morphological characteristics. Principal component analysis reduces the complexity of contextual variables, whilst clustering analysis groups neighbourhoods with similar properties. The analysis reveals four distinct functional neighbourhood types that exhibit consistent patterns across different global cities. The first cluster predominantly features high-entropy neighbourhoods in Singapore and Istanbul with active nightlife and cul-de-sac street designs. The second cluster represents residential areas with morning and midday activity patterns and short-distance mobility. The third cluster comprises larger neighbourhoods with diverse venue types and longer-distance travel patterns, particularly prevalent in Los Angeles. The fourth cluster contains low-diversity areas with afternoon and night activity, characterised by highly walkable street networks. The research demonstrates that similar functional neighbourhood types exist across culturally and geographically diverse cities, suggesting universal patterns in urban human dynamics. These findings provide valuable insights for urban planning and policy-making by revealing how built environment characteristics interact with human behaviour patterns. The framework offers a human-centred approach to understanding cities that complements traditional administrative boundaries with areas defined by actual human interaction and movement patterns. This methodology has significant implications for evidence-based urban planning, enabling planners to identify relationships between built environment features and human behaviours at a global scale.
Key Findings
- Four distinct functional neighbourhood types exist consistently across ten diverse global cities based on human mobility patterns
- High-entropy neighbourhoods with active nightlife predominantly cluster in Singapore and Istanbul, showing cul-de-sac street designs
- Los Angeles shows highest concentration of large neighbourhoods favouring long-distance movements, reflecting car-centric urban structure
- Strong temporal specialisation emerges with neighbourhoods showing distinct activity patterns between midday and night time periods
- The framework successfully identifies functional neighbourhoods that transcend administrative boundaries using human interaction data
Citation
@article{calafiore2021geographic,
author = {Alessia Calafiore; Gregory Palmer; Sam Comber; Daniel Arribas-Bel; Alex Singleton},
title = {A geographic data science framework for the functional and contextual analysis of human dynamics within global cities},
journal = {Computers, Environment and Urban Systems},
year = {2021},
volume = {85},
pages = {101539},
doi = {10.1016/j.compenvurbsys.2020.101539}
}