Research On Semantic Location Network Based On Address Database And Spatial Positioning Method

Nowadays,urban construction has entered the stage of digital city development,and has begun to move towards a smart city.With the advent of the era of big data,great changes have taken place in the ways,methods and sources of data acquisition in urban space.In this context,GIS technology,with its unique ability to manage spatial information,has become an indispensable driving force in the construction of digital city.It has made tremendous progress in fire protection planning,earthquake relief,urban management,municipal management,urban three-dimensional and other fields.At the same time,with the increasing volume of data and the increasing number of domain data,data integration and sharing become more and more difficult.In the long run,the problem of data islands will not disappear easily.How to solve the problem of data fusion,sharing and integration in the development of digital city is a problem that can not be ignored in the construction of digital city.For this problem,many methods have been proposed,but most of them are oriented to data exchange between specific areas,rather than solving the problem of data islands in essence.Because the ultimate goal of digital city is digital earth,and the essence of digital city is to fuse and share all kinds of information according to location,how to realize data exchange between different fields according to location becomes a problem worth studying.Location information takes many forms,including coordinates,postal codes,telephone numbers,IP,place names,addresses and so on.Among these data types,address is relatively standard and rich in usage scenarios.A large number of address data exist in many fields such as transportation,postal,planning,taxation,public security and so on.By locating the address under the urban geospatial system,locationbased data sharing can be effectively completed,which is of great significance to digital cities.However,due to the confusion of China's Toponymic address management system and the significant regional differences of addresses,the address data contained in different departments and domains usually do not belong to the standard address.The existing GIS solutions can not be used to parse these address data,let alone locate them.Aiming at the problem of non-standard address analysis and location in digital city space system,this paper proposes an address coding scheme based on address element structure based on existing address coding solutions.By constructing the relationship between address elements and location,the resolution and location ability of the original address coding solution on non-standard address is enhanced,and assisting the realization of location-based information sharing in digital city.The article mainly contains three parts of research content.The first part is about automatic parsing,recognition and segmentation of nonstandard addresses.Starting from the characteristics of address structure,the method uses statistical features to extract entity names and model features of addresses from the original non-standard address database,and combines rule method to identify and segment address elements.Focusing on this key problem,this paper introduces and analyses the development status of existing Chinese address segmentation methods,discusses the advantages and disadvantages of various methods,and demonstrates in detail the shortcomings of existing methods in resolving non-standard addresses without relying on additional address database.A comparative experiment was set up,and the segmentation effects based on statistics,rules and hybrid methods were analyzed in detail.The effects of the methods on each other and their advantages and disadvantages were also discussed.The experimental results show that the proposed method based on statistics and rules has a good segmentation effect,and can achieve 80% segmentation accuracy,which is significantly higher than the results of the single method.The second part focuses on the standardization process of non-standard address database.On the basis of reviewing the development status of standard address library technology,a standardized process framework of non-standard address Library Based on semantic location network model structure is proposed.In this paper,the problems of address incompleteness,hierarchical relationship error and spatial relationship error in the construction of semantic location network are analyzed,and the specific correction methods are given,and the index construction method used in the construction of semantic location network is given in detail.This paper introduces in detail the semantic relations,spatial relations and mapping methods between semantics and locations contained in the Semantic Location Network(SLN),puts forward the urban address space system based on SLN,and expounds the relationship among address elements,SLN and spatial location.The third part focuses on the location of missing addresses in the address database.According to the semantic relationship contained in the unit nodes of the semantic location network,a location inferring method is proposed.The inferring basis and parameter strategy of the method are discussed.The effects of various strategies on the accuracy of location inferring are analyzed and studied in detail.Experiments on point feature datasets are carried out and compared with standard address matching methods.The experimental results show that the location accuracy of the optimal strategy combination is improved by about 18.34%.