Research On Resident Travel Semantic Method Integrating Multi-source Big Data And Spatio-temporal Weight Model

After combing and analyzing the literature on the space of occupation and residence and the purpose of travel,it is found that most scholars obtain the research data and information by questionnaire,which of the drawbacks includes: 1)requires a lot of manpower,2)takes up a long time,3)the number of samples obtained is limited,4)lack of representativeness and typicality,5)the problem of too single data source affects the accuracy of the research results.At present,a series of advanced technologies,such as perceptual positioning technology and information technology,are becoming more and more mature,and have been popularized and actively applied in many fields.The data related to residents' travel time and space trajectory have also become more and more perfect,which provides an important basis for people to analyze urban space and understand residents' travel mode accurately.Based on the semantic model of spatio-temporal weight travel,combined cell phone signaling data and POI data with characteristics of mobile phone signaling data and poi data,spatial recognition method and the POI quantitative recognition method are constructed respectively in Guiyang.These methods can be used to analyze the characteristics of population distribution,occupation housing,travel OD,construct semantic analysis method of resident travel combining multi-source big data and spatiotemporal weight model,dig out the rules of user behavior,study the spatial form of urban occupation and residence in detail and accurately grasp the distribution of hot spots.The main contributions of the paper are as follows:(1)The signaling data of mobile phone is used in the investigation and analysis of the working and living space in the city.This paper is based on the previous research results,further discussion and analysis are made through the division algorithm of OD points,based on the above,a new method of urban job-occupancy space identification is proposed.Mobile Phone Signaling Data as a New Track Data of Urban Residents' Travel Information in Internet Age to provide a new perspective for urban dynamic space research.Based on the mechanism of mobile phone location data formation,this paper combs and constructs a number of mobile phone data preprocessing algorithms,such as invalid data filtering,drift and so on,which can preprocess the data quickly and efficiently.It is easier to obtain accurate and objective resident location data.(2)Construction of spatiotemporal weight model based on POI data.City POI data reflect the static spatial distribution of urban geographical entities,by studying the laws of time and space visited by various POI,this paper studies,Weight Model of Residents' Travel Purpose in Deep Analysis,on the basis of kernel density calculation,the heat map is used to describe the agglomeration of various POI,and the frequency and weight of all kinds of POI data are counted every hour of the day,which provides an important basis for the purpose of mining residents' travel in depth.(3)The semantic position solving method based on trajectory mining is defined.There are many limitations in existing trajectory mining,It is difficult to collect user semantic information accurately and efficiently,to determine the user's position state scientifically and reasonably,and to estimate its purpose and requirement accurately and reasonably.It is impossible to solve the problem of information adaptation demand of complex location perception application efficiently and accurately.After extensive and in-depth study,this paper proposed a semantic location calculation method based on trajectory mining and constructed a framework system of trajectory data mining with wide applicability.The study combines the advantages of multi-source data to accurately identify the spatial distribution of urban occupation and residence and identify the purpose of residents' travel,and provides an important basis for urban planning and traffic management.