Abstract
In this study, we analyze urban traffic flow using taxi trajectory data to understand the characteristics of traffic flow from the network centrality perspective at point (intersection), line (road), and area (community) granularities. The entire analysis process comprises three steps. The first step utilizes the taxi trajectory data to evaluate traffic flow at different granularities. Second, the centrality indices are calculated based on research units at different granularities. Third, correlation analysis between the centrality indices and corresponding urban traffic flow is performed. Experimental results indicate that urbaxperimental results indicate that urbaxperimental results indicate that urban traffic flow is relatively influenced by the road network structure. However, urban traffic flow also depends on the research unit size. Traditional centralities and traffic flow exhibit a low correlation at point granularity but exhibit a high correlation at line and area granularities. Furthermore, the conclusions of this study reflect the universality of the modifiable areal unit problem.
Highlights
The rapid development of GNSS and communication technology has resulted in the emergence of large amounts of GPS trajectory data, thereby facilitating the analysis of urban dynamics (Ahas et al, 2010; Jia and Jiang, 2012; Kang et al, 2014)
Betweenness centrality is utilized to characterize urban traffic flow, and the results demonstrate that the traditional betweenness centrality is unsuitable for analyzing the dynamic process of traffic flow and needs to be further improved (Kazerani and Winter, 2006a)
We propose a method of urban traffic flow analysis at different granularities on the basis of taxi trajectory data
Summary
The rapid development of GNSS and communication technology has resulted in the emergence of large amounts of GPS trajectory data, thereby facilitating the analysis of urban dynamics (Ahas et al, 2010; Jia and Jiang, 2012; Kang et al, 2014). Numerous studies have extensively analyzed urban traffic flow from the road network perspective. Network centrality is an important indicator of road network characteristics and has been broadly used in the analysis of urban traffic flow. Betweenness centrality is utilized to characterize urban traffic flow, and the results demonstrate that the traditional betweenness centrality is unsuitable for analyzing the dynamic process of traffic flow and needs to be further improved (Kazerani and Winter, 2006a). Gao et al (2013) maximized taxi trajectory data in evaluating actual traffic volumes and analyzed the feasibility of predicting urban traffic flow using betweenness centrality. The traditional betweenness centrality is not a good predictor because it disregards the distance decay effect
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
