Research On Data Quality Assessment Of Multi-source Heterogeneous Location Data

With the continuous acceleration of the informationization process of the society,the continuous maturity of the sensing technology and computing environment,and the continuous improvement of people's living standards,positioning devices such as civilian GPS are widely used in vehicles and mobile terminals.Therefore,there is a rapid growth with location-based big data,and multi-source location data composed of geographic data,vehicle GPS trajectory,mobile phone positioning data and user"check-in" records has become an important strategic resource for sensing the laws of human community activities and building smart cities.However,the current location data has the characteristics of wide source,various types,various forms of expression,fast update,and large amount of data.The low quality problems cannot be ignored,because the location data mainly comes from various types of sensors,video surveillance,mobile terminals,and floating.Car GPS system,etc.,data acquisition and transmission process is susceptible to a series of factors such as signal propagation,mobile terminal equipment,weather,etc.The collected data will produce many quality problems such as data loss,invalidity,error,ping-pong phenomenon,drift phenomenon and so on.However,the research on the quality of existing location data is mostly directed to the study of single source data such as GPS trajectory data.There are few studies on the quality of location data from other sources,and the evaluation methods are relatively simple,and the analytic hierarchy process(AHP)is often used.In this paper,the quality of multi-source location data is taken as the research object,and the quality of the multi-source location data is evaluated systematically.First,multi-source data collection,including taxi GPS trajectory data,Sina Weibo check-in data,associated POI location data,and mobile handset location data.Secondly,it analyzes the problems existing in multi-source location data,comprehensively considers the application requirements of each location data,proposes corresponding evaluation indicators,and constructs a general multi-source location data quality measurement framework.Thirdly,an evaluation model for each quality indicator is determined,and a comprehensive evaluation method based on the combination of the G1 method and the anti-entropy weight method is proposed.Finally,the multi-source location data quality measurement framework proposed by us is applied to the quality assessment of the actual multi-source location data,and the quality level of the collected multi-source location data is analyzed.The feasibility of the quality measurement framework and assessment method proposed in this paper is verified.The quality measurement framework proposed in this paper takes into account the various quality problems faced in the location data,and has applicability,which provides a reference for the quality research of subsequent location data.At the same time,the research of this paper also expands the field of big data quality research,which has good theoretical significance and practical value.