How friends share urban space: An exploratory spatiotemporal analysis using mobile phone data

Yang Xu, Alexander Belyi, Iva Bojic, Carlo Ratti

Research output: Journal article publicationJournal articleAcademic researchpeer-review

20 Citations (Scopus)


The framework is applied to a Call Detail Record (CDR) dataset collected in Singapore. First, a comparative analysis is performed to understand how friends share urban space differently from random people. Then, we derive two metrics to quantify the “bonding” and “bridging” capabilities of places in the city. The two metrics reflect the potential of a place in connecting friends and random people (e.g., chance encounters), respectively. Finally, we examine the temporal signature of the places’ bonding capabilities, and associate the results with various types of Points of Interest (POIs). We find that: (1) friends are more likely to share urban space than random people, and they also share more locations; (2) a place could play different roles in connecting friends vs. random people, and the relationship (between bonding and bridging) varies depending on the time and type of a day (weekdays vs. weekends); (3) the temporal signature of bonding capability is strongly related to the semantics of a place; (4) certain POI types (e.g., shopping malls) tend to have a much higher impact on bonding capability than others (e.g., sports centers).
Original languageEnglish
Pages (from-to)468-487
Number of pages20
JournalTransactions in GIS
Issue number3
Publication statusPublished - 1 Jun 2017
Externally publishedYes

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)

Cite this