点状地名信息的加权泰森多边形检索法  

A Weighted Voronoi Diagram-based Retrieval Approach for Point-like Toponym Information

作　　者：张宇[1,2,3] 王琦 吴文周[1] 苏奋振 ZHANG Yu;WANG Qi;WU Wenzhou;SU Fenzhen(State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,CAS,Beijing 100101,China;University of Chinese Academy of Sciences,Beijing 100049,China;Beijing Normal University,Beijing 100875,China)

机构地区：[1]中国科学院地理科学与资源研究所,北京100101 [2]中国科学院大学,北京100049 [3]北京师范大学,北京100875

出　　处：《测绘学报》2017年第11期1919-1926,共8页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金创新研究群体科学基金(41421001)~~

摘　　要：鉴于地名库中多数地名仅以中心点坐标形式记录其空间位置信息,缺乏其空间范围的具体描述,造成地名检索中的片面性和局限性,本文在深入分析地名及其空间属性、空间关系的基础上,考虑同类型地名的不同性质在检索中的重要作用,利用泰森多边形在地名边界近似中的优势,提出了针对点状地名信息的加权泰森多边形检索法,依据同类型地名的面积属性作为权重指标构建泰森多边形,以近似表达地名的空间范围,进而描述地名间的各类空间关系;并从拓扑关系、方位关系与距离关系3个方面给出了具体公式以计算地名间的空间相似性;最后以行政区划边界近似与检索为例验证该方法。试验结果表明,该方法可较好地近似表达地名空间范围及空间关系,增强了检索词与地理信息资源在空间位置上的相似性度量,检索结果优于传统方法。Considering the fact that most of the place names in the gazetteer record the spatial information only in the form of center coordinates,which lacks the description of the specific spatial extent of the place name,one sidedness and limitations exist in the toponym information retrieval.Based on the fully analysis of the place names,spatial attributes and spatial relations,the paper proposes a weighted Voronoi diagram based retrieval approach for point like toponym information,which considers the importance of different properties of the place names belonging to the same type in the information retrieval,and takes advantage of the weighted Voronoi diagram in the boundary approximation of the place.The approach constructed weighted Voronoi diagram based on the area properties of the same type place names,and approximated the spatial extent of the place.Then,spatial relation between two places could be described,and the spatial similarity was achieved by the given equations to calculate the similarities of the topological,directional and distance relationship.The approach was verified by approximating boundaries of administrative divisions and retrieving their information.The experiment results show that the approach can be better applied in the approximation of spatial extents and relations of place names,and the spatial similarity matching between query words and geographical information resources can be improved.Finally,the proposed approach is better than the traditional approaches.

关 键 词：加权泰森多边形 地名信息检索 地名 边界近似 空间关系相似性 

分 类 号：P208[天文地球—地图制图学与地理信息工程]