Exploring The Characteristics Of Urban Taxi Traveling Behaviors Based On Network Motifs

With the popularization of vehicle-mounted GPS positioning devices,large-scale and high-quality spatial-temporal dataset of urban taxi becomes accessible.Taxi trajectory data records the traveling information of taxis and the spatiotemporal information of the OD points of passengers,which makes it gradually become a reliable data source in research fields such as analysis of urban residents' traveling behavior,urban traffic situation management and urban spatial structure optimization.On one hand,the analysis and exploration of taxi travel characteristics are of great significance for taxi dispatch,taxi route selection,and optimization of urban public transportation services.On the other hand,it provides an important basis for understanding the laws of dynamic space-time evolution of individuals,groups,and cities.This paper introduces the theory of complex network and the concept of network motif into the process of taxi travel feature exploration.By identifying the network motifs of the taxi OD flow network,the spatiotemporal analysis of the taxi traveling feature based on the network motifs is performed from a new research perspective.This research aims to explore the overall and local laws of taxi trips by analyzing the basic structural units of the OD flow network.Based on network motif identification,the characteristics of taxi trips are explored by analyzing the structure,functional and statistical distribution characteristics of the motifs identified.On this basis,the time dimension is introduced,and the network motif recognition algorithm is performed on the data of different date types and different time periods to compare and analyze the spatial and temporal distribution characteristics of taxi trips.The main contents and results of this paper can be concluded as follows.(1)Network motif recognition of OD flow network.Based on the research area and experimental dataset,the directed OD flow geographic network is constructed by applying the urban grids as the nodes and the OD flow as the directed edges.Besides,several corresponding random networks are generated.For the above networks,a subgraph identification algorithm is performed based on the enumeration and common substructure,which uses the adjacency matrix classification as the standard for subgraph isomorphism.The significance test of the subgraph recognition results is carried out to obtain network motifs with a frequency that is much higher than that of the random networks.This part lays the foundation for the experimental data for the analysis of taxi trip characteristics.(2)Analysis of the statistical characteristics of taxi trips based on network motifs.In this research,the statistical characteristics of taxi trips are analyzed based on the network motifs identified above.Based on the structural characteristics of the network motifs and its composition and statistical distribution characteristics in a complex network,the analysis is carried out from three aspects,which are the whole,multiple date types and multiple time periods.(3)Analysis of the spatiotemporal characteristics of taxi trips based on network motifs.This part also adds the temporal information into the analysis procedure.The network motifs are recognized from multiple date types(weekdays and weekends)and multiple time periods.By comparing the difference of motifs in quantity and spatial distribution,the characteristics of the temporal and spatial changes of taxi trips are analyzed and summarized.Also,the changes are understood by analyzing different needs of urban residents in different time periods and different spatial regions.