Research On Key Technology Of Semantic Consistency Processing For Multi-sources Vector Data

The rapid development of national economy and the need of national defense construction have accelerated the building of geographic information database,which makes the application scope of geographic information data expand and the application level of geographic information data deepen continuously.Geographic information data with better current situation and consistent semantic information is urgently needed by various fields and industries.Meanwhile,due to the different application requirements,geographic information data production departments work relative independently,which leads that the semantic inconsistency of multi-source vector data is becoming increasingly prominent,and seriously affects the information exchange and sharing.Eliminating semantic heterogeneity of multi-source vector data is not only the main task of establishing a unified geographic information data sets,but also the basic premise of providing geographic information data services.Therefore,there is an urgent need for a set of theories,methods and technologies to eliminate the semantic heterogeneity of multi-source vector data,provide technical support for the establishment of consistent semantic vector data sets and satisfy the urgent needs of users for high-quality geographic information data.This dissertation focuses on the key technologies of semantic consistency processing of multi-source vector data,aiming to solve the semantic inconsistency problem in multi-source vector data,in order to provide theoretical and technical support for establishing the geographic information data sets with consistent semantic and higher precision.The main work completed and the results obtained of this dissertation are as follows:1.Combined with the practical application requirements,the research background and significance of this dissertation are analyzed.Related to the semantic consistency processing of multi-source vector data,the research status at home and abroad about geospatial data consistency,geographic classification semantics,and identical geographic entity matching are summarized.The shortcomings existing in current research are pointed out,then the research objectives and content of this dissertation is proposed.2.The definition of semantic consistency of multi-source vector data is clarified,and the causes and specific manifestations of semantic inconsistency are analyzed.Combined with the dual elaboration of geographical ontology on philosophical ontology and information ontology,the semantic structure of vector data is discussed from the semantic level of vector data expression,and the strategy and technical process of multi-source vector data semantic consistency processing is put forward.3.The semantic consistency processing of geographic categories is studied.Basic methods and coding principles of geographic features classification are summarized,and the semantic relationship between geographic categories is analyzed.The categories mapping method from different perspectives of classification semantic understanding in this dissertation.The semantic understanding mechanism of geographic features classification based on descriptive knowledge is expounded,and a geographic categories mapping method considering descriptive knowledge is proposed which can establish the corresponding relationship of geographic categories by measuring the comprehensive semantic similarity.Ontology attribute characteristics of geographic categories is summed up,and the general methods of ontology attribute characteristics extraction and vectorization expression are established.Then a geographic categories mapping method based on ontology attribute characteristics learning is proposed to realize the mapping of geographic categories.4.The matching processing of identical geographic entity is studied.The basic idea of identical geographic entity matching is elaborated.Aiming at the problem of geographic entity matching based on semantic characteristics,a identical geographic entity matching method with the constraint of multiple semantic attributes is proposed,after designing the different attribute similarity measure algorithms and attribute weight determination methods,to realize the identical road entities matching.In the geographic entity matching based on geometric characteristics,fréchet distance is easily affected by the curve vertices distribution and the sampling accuracy when uesd to measurse the distance of identical linear features.A identical geographic entity matching method with the improved fréchet distance by using facing project is proposed which can enhance the matching quality of linear geographic entities.5.The semantic consistency processing of geographic feature attributes is studied.The basic content of attributes mapping and transformation of multi-source vector data is analyzed.Then the transformation rules based on production structure are designed.XML and XML schema are used to classify,describe,store and manage the production attribute transformation rules,which are realized by the corresponding rule template.Finally,a attributes consistency processing method based on rule control files is established,and an example is used to analyze the process.6.A prototype system for semantic consistency processing of multi-source vector data is developed.The main functions of the system and the related experimental data are introduced.The key technologies and methods proposed in this dissertation are further verified by experiments.