Evaluating transit-served areas with non-traditional data: An exploratory study of Shenzhen, China

Abstract

In this study, transit-served areas (TSAs) are defined as areas within a reasonable distance (e.g., 800 meters) of transit services. TSAs have two key dimensions: physical features (e.g., land-use density and mix) and performance (regarding human behaviors). Non-traditional data (NTD) (e.g., social media check-ins and cellular network data) can supplement traditional data (TD) (e.g., interviews and censuses) to enhance studies and monitoring of TSAs. A case study of Shenzhen, China, illustrates how to combine NTD and TD to evaluate the features and performance of 167 TSAs along metro lines. It finds that NTD can be used to formulate new indicators to measure and monitor the two dimensions of TSAs; the features and performance of different TSAs vary significantly; point of interest (POI) efficiency, or the average users attracted by each POI, can be a useful indicator to differentiate TSAs’ performance; the POI efficiency of a single TSA can vary across days and the POI efficiency of an extremely efficient or inefficient TSA can be totally different across days; and the combination of NTD and TD can effectively help locate extreme TSAs and explain factors contributing to the extremity.