Research Of Geographic Entity Matching Technology

With the rapid development of information technology and continuously proposals of new requirements, the significance of spatial data is increasing; the range of application is becoming more and more extensive. Due to the differences among time, scale, collectors and methodologies of data collection, obtained spatial data are divergent in geometric shape, precision and specification, which results in a great deal of inconveniences. As the essential technology of spatial data integration and sharing, identical entity matching technology has twofold aims, establishing mapping relation of identical elements based on a multi-source geographic database whilst accomplishing an integrative use of multi-source data. This thesis attempts to investigate geometric, topological and semantic matching of identical entities through theoretical and mathematical approaches. The main contents and findings are listed below:(1) Discussing basic concepts, characteristics, difference and similarity of spatial data which provides theoretical framework for identical entity matching. Thus, analyzing definition, constraints, technical process and algorithmic assessing criterion of identical entity are discussed.(2) Investigating matching methods from three perspectives, geometric, topologic and semantic matching. This paper also looks into the matching standard for point entity, polyline entity and polygon entity, based on measurement of similarity of spatial data. The standard contains five components i.e. length, distance, area, shape, and direction. It also presents appropriate methods of describing characteristics of geometric entity.(3) Constructing a further study of relevant strategies and methods employed in the matching process, which includes using bidirectional matching strategies to provide solutions for the non 1:1 situations and improve the rate of accuracy; generalizing the assessing measurements, analyzing applying conditions and characteristics of each entity, elaborating the comprehensive matching methods of point entity, polyline entity and polygon entity separately; establishing a quick access to candidate matching sets, excluding impossibly matched entities so that the overall effectiveness can be improved. This section discusses three acquiring methodologies based on buffer zone, the minimum bounding rectangle and spatial index.(4) Under the Windows circumstance, the experimental system is based on the C # software developmental platform, utilizing COMGIS (component GIS) technology, ArcGIS Engine 10 components to exploit visualization of maps and frequently-used functions of GIS. Operationalizing this kind of system contributes to accomplishing some significant functions such as data pretreatment and matching of point entity, polyline entity and polygon entity. The system validates and analyzes the matching methods and strategies.