The Construction And Visual Analytics Of Spatiotemporal Knowledge Graph Towards Urban Traffic

Urban traffic is an important part of social life,which affects the operation of the city and people’s daily life.With the advancement of urbanization,the rising urban population and the strong investment in urban infrastructure,the content of urban traffic is becoming richer,the coupling is closer and the relationship is more complex.It is a challenge to the comprehensive expression and efficient understanding of urban traffic.Urban traffic knowledge is the result of human cognition of the changing law of spatial distribution and correlation caused by the movement of things in traffic scenes.Exploring how to express urban traffic knowledge and digital twin urban traffic to help people deepen their comprehensive perception and understanding of urban traffic has become an urgent problem to be solved in the development of intelligent transportation and smart city.As a replica of urban travel activities,trajectory data contains information about when,where and how people move in the city,and provides a rich source of knowledge for urban traffic expression.The rich graphical expressions of computers improve people’s ability to perceive complex knowledge.Based on the background above,this paper takes the urban traffic with rich connotation as the research object,the trajectory data as the main data source,and the perception and understanding of urban traffic knowledge as the goal to develop urban traffic-oriented knowledge model,study urban traffic knowledge generation method based on trajectory data,and efficient perception technology of urban traffic knowledge.The work of this dissertation is carried out in the following four aspects:1.In order to solve the problem that urban traffic knowledge is rich in content,complex in relationship and difficult to express,an urban traffic knowledge representation model based on spatiotemporal knowledge graph is proposed.Urban traffic describes the dynamic process of group movement and the change of spatial distribution and relationship of things caused by group movement,and describes urban traffic difficulties comprehensively and systematically.Combined with the advantages of knowledge structure and relevance of knowledge graph,and starting with the characteristics of time,space and dynamic,this paper puts forward a spatiotemporal knowledge graph considering time and space.The spatiotemporal knowledge graph is used to construct the entity,attribute and relationship knowledge model for urban traffic,and a structured urban traffic knowledge base is formed.2.In order to solve the problem of the explosion of urban trajectory data and the lack of traffic knowledge,a framework of urban traffic knowledge generation based on trajectory data is proposed.Trajectory data contains rich semantic knowledge and is an effective source of urban traffic knowledge.This paper is oriented to the urban traffic spatiotemporal knowledge graph,an urban traffic knowledge generation system is proposed,which includes traffic knowledge extraction considering the semantics of trajectory data,attribute knowledge generation with time-varying characteristics,spatial dynamic entity knowledge generation of urban traffic events and multiple relationship generation for traffic scenes.3.In order to solve the problem that the content of urban traffic spatiotemporal knowledge graph is huge and difficult to perceive,a visual analytics method for urban traffic spatiotemporal knowledge graph is proposed.In order to meet the”customized knowledge” and ”graphical expression” demand of urban traffic knowledge,a visual analytics framework is constructed,which includes visual interaction,knowledge acquisition and visual expression.in this paper,various forms of humancomputer interaction methods are proposed,and a high expressive visualization method based on entity spatial distribution,dynamic change and association relationship in traffic spatiotemporal knowledge graph is designed.4.Using multi-source static spatial data and trajectory data collected in Shenzhen,multiple traffic entities,attributes and relational knowledge are generated,and Shenzhen traffic spatio-temporal knowledge graph is constructed based on the urban traffic knowledge graph model proposed in this paper.A visual analytics system for efficient perception of urban traffic is designed.Taking the urban crowd aggregation analysis as an example,it is verified that the multivariate visualization method in the visual analysis system is helpful to knowledge perception.The traffic congestion analysis is taken as an example to verify that the structured knowledge organization of spatiotemporal knowledge graph is helpful to urban traffic knowledge discovery.A questionnaire survey is applied to evaluate the performance of proposed methods on urban traffic knowledge perception by comparing with an open urban traffic visualization platform and an existing visual analytics design.The score results show that the proposed method performs better in visualization design and ability to perception the visualized knowledge.