Between April and July 2019 we participated to the Future Cities Challenge organized by Foursquare and an international team of scientists. Big data about human mobility was provided for 10 megacities around the world.
Increasing evidence supports the fact that cities are complex systems, with a broad spectrum of structural and dynamical features which lead to unexpected and emerging phenomena. Understanding urban dynamics at individual level -- but also as the outcome of collective human behaviour -- will open the doors to uncountable applications ranging from enhancing the sustainability and the resilience of the city to improving health and well-being of its inhabitants.
Here, we use a unique data set of longitudinal human flows provided by Foursquare, a leader platform for location intelligence, to characterize the functional organization of a city. First, we build multidimensional network models of human flows corresponding to different types of activities across time. We quantify the efficiency of flow exchange between areas of a city in terms of integration and segregation, respectively. Results reveal unexpected complex spatio-temporal patterns that allow us to gain new insight on the function of 10 megacities worldwide. We discover that large cities tend to be more segregated and less integrated, and that human flows at different hours of the day or between different types of activities enable the identification of different ``cities within the city'' which indeed show clear dissimilarities in terms of both functional integration and segregation.
Our analysis provides new insights on how human behaviour influences, and is influenced by, the urban environment and, as an interesting byproduct, to characterize functional (dis)similarities of different metropolitan areas, countries, and cultures.
Modeling Structure and Function of Urban Systems. Left: Urban structural backbone of the 10 megacities considered here, as described from their street networks (data obtained from Open Street Map. Middle: Urban functional networks described by the Foursquare data. The nodes are obtained by dividing the area analysed into cells of 500m x 500m. The edges are subsequent check-ins that might be between activities of the same type (intra-links: e.g. Food-Food, Tourism-Tourism) or different types (inter-links: e.g. Food-Tourism, Food-Sport). The collection of layers and inter-layer flows defines a multilayer network, i.e., a multidimensional functional representation of the urban areas. Right: The mobility flows between areas are captured as the edges' weights. In the example, describing New York City, we can observe the different spatial distribution of flows between and across different activity layers.
Disentangling human flows. We illustrate here, for four of the ten cities studied in this paper, the strikingly distinct views on the functional organization of a city extracted by isolating intra- or inter-layer flows. These maps outline the different cities within the city which we disentangle by decoupling the urban flows into activity-aware multilayer networks.
Human Flows Data
The Foursquare data made available for the Future Cities Challenge describe 24 months of check-ins between April 2017 and March 2019 (included). The 10 world mega-cities included in the challenge are Chicago, Istanbul, Jakarta, London, Los Angeles, Tokyo, Paris, Seoul, Singapore and New York City (see Figure). 104,657,168 checkins from 331,264 venues are considered.
The flows between different areas are derived by subsequent check-ins to the Foursquare's location-based services and coarse grained with a 500m x 500m granularity. In the data provided, check-ins are already aggregated by couple of venues (origin and destination), month and hour of the day (morning, midday, afternoon, night, and overnight).
The metadata of the venues include a category field which describes the type of venue in great detail (e.g.: Knitting Stores, Mini Golf Courses, Rock Clubs, ...). We defined a set of macro-categories we used to define a limited number of layers.
By disentangling the mobility flows into a multilayer network structure, we are able to quantify the differences in the functional organization of the different cities within a city that are outlined by movement between different types of activities in a limited number of layers.
City size, fraction of hotspots, and functional segregation. Left: As the size L of the city grows, the fraction of area that is represented by hotspots obtained with the LouBar method decreases (Pearson correlation coefficient of -0.65). Right: Having in proportion a larger fraction of the urban area covered by hotspots makes the cities less segregated (Pearson correlation coefficient of -0.61), and at the same time the more integrated.