Study On Semantic Enhancement Of Trajectory Data For Smart Traveling

The smart city is an advanced concept of urban development and has received wide attention.The smart travel is an important component of the smart city.It not only optimizes urban transport planning,but also improves people's travel efficiency.The rise of smart travel has changed the traditional mode of travel and promoted the intelligence of traffic travel,achieving the purpose of alleviating traffic congestion and optimizing the spatial and temporal distribution of travel.In the field of smart travel,the widespread use of GPS enables us to obtain a large amount of travel trajectory data at a low cost.These trajectory data contain rich semantic information,such as road network topology,people's travel purpose,and so on.Further,the road network topology can be used to build a digital city map.The trajectory data contains rich semantic information and make itself have deep mining and enhancement significance.In order to deeply mine and enhance the semantic information contained in the trajectory data,this paper mainly did the following work:In order to obtain accurate digital map,this paper proposes a three-phase urban digital map mining framework.The framework firstly extracts roads based on the distribution of travel trajectories on the road.Secondly,based on the extracted roads,the framework uses the SIFT algorithm to determine the intersection locations and its types.Finally,with prior knowledge,the framework combines the intersections and roads found in the trajectory data into a digital map.In the experimental part,this paper uses trajectory data from the DiDiChuXing company to verify and evaluate the effectiveness of the framework.In order to fully understand people's travel purpose,this paper propose a probabilistic two-phase framework named TripImputor,for making the real-time taxi trip purpose imputation and recommending services to passengers at their drop-off points.Specifically,in the first phase,we propose a two-stage clustering algorithm to identify candidate activity areas(CAAs)in the urban space.Then,we extract fine-granularity spatial and temporal patterns of human behaviors inside the CAAs from foursquare check-in data to approximate the priori probability for each activity,and compute the posterior probabilities(i.e.,infer the trip purposes)using Bayes' theorem.In the second phase,we take a sophisticated procedure that clusters historical drop-off points and matches the drop-off clusters and CAAs to immerse the real-time response.Finally,we evaluate the effectiveness and efficiency of the proposed two phase framework using real world data sets,which consist of road network,check-in data generated by over 38000 users in one year,and the large-scale taxi trip data generated by over 19000 taxis in a month in Manhattan,New York City,USA.In order to accurately label the trajectory semantic,this paper proposes a semantic labeling model with two phases.The model first converts the travel trajectory into a travel trajectory graph.The graph contains not only the starting point of the trip but also the physical space around the trip.Second,the model will use the trajectory graph as an input of the graph convolution neural network.Further,this paper uses the graph convolution network to learn the characteristics of travel.In the end,this paper inputs features into a fully connected network for semantic labeling.In the experimental part,this paper uses the data from New York to design and optimize the convolution neural network model framework of this paper.At the end of the paper,the main work of the full text is summarized and prospected,and some deficiencies in the work are proposed.