Tracking the evolution of temporal patterns of usage in bicycle-Sharing systems using nonnegative matrix factorization on multiple sliding windows. Issue 2 (3rd April 2018)

• Record Type: Journal Article
• Authors: Cazabet, Remy; Jensen, Pablo; Borgnat, Pierre
• Is Part Of: International journal of urban sciences. Volume 22:Issue 2(2018)
• Page Start: 147
• Page End: 161
• Publication Date: 2018-04-03
• Subjects: Nonnegative matrix factorization -- bicycle sharing system -- activity-based modelling -- temporal usage patterns; Cities and towns -- Pacific Area -- Periodicals; Cities and towns -- East Asia -- Periodicals; Urban ecology (Sociology) -- Pacific Area -- Periodicals; Urban ecology (Sociology) -- East Asia -- Periodicals; Cities and towns; Urban ecology (Sociology); Pacific Area; Electronic journals; Periodicals; 307.7609505
• DOI: 10.1080/12265934.2017.1336468
• Languages: English
• ISSNs: 2161-6779

Abstract

ABSTRACT: Bicycle-Sharing Systems (BSS) are growing quickly in popularity all over the world. In this article, we propose a method based on Nonnegative Matrix Factorization to study the typical temporal patterns of usage of the BSS of Lyon, France, by studying logs of rentals. First, we show how this approach allows us to understand the spatial and temporal usage of the system. Second, we show how we can track the evolution of these temporal patterns over several years, and how this information can be used to better understand the BSS, but also changes in the city itself, by considering the stations as social sensors.