Examining the varying influences of built environment on bike-sharing commuting: Empirical evidence from Shanghai

Journal article, 2022

Commute behaviors, as the primary part of urban mobility, remains largely underexplored, especially for bike-sharing users. Recent development in data availability open up new possibilities to delve into bike-sharing commuting over long-term periods on a large scale. This study proposes a methodological framework that enables a logical identification of bike-sharing commuting activities and a comprehensive examination of urban built environment effects on shaping commuting patterns. To this end, a series of data mining methods are developed in support of the identification of regular bike-sharing commuting, and the concepts of home-work balance and mobility trend are proposed to describe underlying commuting patterns. The XGBoost model and Necessary Condition Analysis (NCA) method are then adopted respectively to test the sufficiency and necessity of built environment on commuting patterns. The results confirm the massive existence of individual-level bike-sharing commuting activities and the pivotal role of bike-sharing in urban commuting. Also, the spatial distributions of home-work balance and mobility trend driven by job-housing separation show different clustering patterns. Besides, the synergy of sufficiency analysis and necessity analysis investigates the complex interplay of built environment-commuting patterns. This critical analysis of bike-sharing commute provides insights into sustainable transit planning and urban design.