Authors: Shaojun Liu*, Nanjing Normal University
Topics: Geographic Information Science and Systems, Social Geography, Urban Geography
Keywords: smart card data, temporal network, spatiotemporal movement pattern, centrality measure, evolutionary clustering
Session Type: Paper
Start / End Time: 10:00 AM / 11:40 AM
Room: Bayside A, Sheraton, 4th Floor
Presentation File: No File Uploaded
The development of location-aware technology has led to the generation and enrichment of multi-source location data, which has been widely used in some fields such as ubiquitous mapping, intelligent transportation and smart city. Smart card data records personal travel information of public transport travelers in the city, which can reflect urban human mobility patterns and characteristics. There has been a great deal of research on the individual and collective mobility patterns, but the research on the dynamic evolution is deficient. This paper attempts to propose a mining framework based on smart card data to explore the spatiotemporal evolution pattern of urban human flows. Based on the theory and practice of dynamic network analysis, the framework includes three parts: (1) aggregating the individual traffic card records into time-tagged OD flow dataset; (2) the evolution pattern of the urban concentration and diffusion center could be analyzed by calculating two temporal centrality metrics; (3) the evolution of community structure would be analyzed by using evolutionary clustering method. Finally, we present a case study with a large subway dataset of smart card data in Shanghai, China to demonstrate and evaluate this framework. The mining of urban human spatiotemporal mobility evolution patterns makes an important progress to the study of urban mobility patterns: on the one hand, it deepens the understanding of the formation, development and evolution mechanism of urban movement patterns; on the other, the analysis of temporal and spatial mobility patterns is the basis for the prediction of urban mobility geospatial contacts.