Research On GPS Trajectory Modeling And Semantic Enhancement For Users' Travel Characteristics

With the continuous development of our country's society and the continuous advancement of urbanization,people face enormous tests and challenges In traffic,environment,housing etc.Traffic congestion,unreasonable route planning and other issues are increasingly severe,which constrained the city' s sustainable development.Residents' daily travel is an important part of urban transport and one of the main sources of traffic pressure.Therefore,studying the characteristics of residents' trips,understanding the residents' travel purposes,travel time,and travel modes will help to.further understand the characteristics of urban transport needs.This will provide data support for design and planning of transportation and strategic long-term support.In recent years,with the development of position sensing devices and related technologies,devices that can be used to record location information have become more and more popular.Vehicles and mobile terminals such as smart phones can be directly installed with GPS receivers or embedded GPS chips.Users can easily obtain personal location information,use various location-based services(Location Based Service-LBS),and record their own movements in the form of trajectories.GPS trajectory data is of great significance for revealing the behavioral characteristics and laws of users.However,the current research of this kind of GPS mobile data is mainly limited to analyzing the original trajectory generated by the motion sensor without fully considering the related higher level semantics.This paper further explores the semantic information of GPS trajectory moving objects for user-oriented travel feature applications.For this purpose,this paper proposes a general GPS trajectory modeling,segmentation and semantic annotation method.Due to structural and semantic heterogeneity,modeling and semantic enhancement of trajectory data are still challenging.This paper aims to reduce the structural heterogeneity among data sources and specify the semantics of the concepts in a definite way,which has certain advantages for solving the above problems.Therefore,this paper first proposes an ontology-based semantic trajectory model that covers common structures encountered in the trajectories associated with applications and geographic module links.GPS trajectory ontology is divided into geometric trajectory ontology and semantic trajectory ontology.Geometry trajectory ontology describes the time-space features of the composition trajectory,and includes the concept of the geometric description of GPS trajectory.It is an indispensable part of semantic trajectory modeling.Semantic ontology includes the description of geographic location information and application domain information.Secondly,based on the construction of trajectory ontology,this paper designs and proposes a method for determining the stopping point of GPS trajectory,which achieves the acquisition of the stop point of GPS trajectory and the segmentation of geometric trajectory.The trajectory was first upgraded to a "structured trajectory";A semantic annotation framework based on GPS structured trajectories and related semantic annotation methods were finally proposed.Through a series of calculations,the GPS trajectory was upgraded to a semantic trajectory with rich semantic information."in order to provide a higher level of information support for the analysis and application of user travel characteristics."...