Identifying spatial structure of travel modes through community detection method

Abstract

Identifying the spatial structure of travel activities could help better understand the urban structure, as the structure of a city is heavily shaped by its transportation system. Today the emerging traffic big data contributes a lot to delineate daily travel activity of individuals. In this paper, multi sources of travel data are utilized to give a full description for the spatial structure of travel modes in a city. Firstly, the massive intra-city travel flow data are constructed as connection embedded networks. Then, using the community detection method and the network analysis method, the latent spatial connection structure of travel modes and corresponding properties of sub-networks are investigated. Finally, a case study of Guangzhou is conducted using travel data of three travel modes. Results imply that revealed structures correlate with the land use and the geographical distance. The taxi flow network shares similar characteristics with the bus flow network, but the mass transit railway system is more independent.