摘要
针对目前大多数空间聚类算法主要是针对关系数据且没有考虑空间拓扑关系相似性的问题,对基于空间拓扑关系的空间聚类方法进行研究.提出了一种基于线面包含关系的GML(geography markup language)空间聚类算法SCGML-LRI(spatial clustering in GML data based on line-region inclusion relations).算法将GML文档中线面空间对象的包含关系作为空间对象相似性度量准则,并用CLOPE算法对空间对象进行聚类.实验结果表明:算法SCGML-LRI能实现GML数据的空间聚类,并具有较高的效率.
For solving the problem most spatial clustering algorithms deal with the relational data without consideration of the similarity of spatial topological relations. A method for spatial clustering based on spatial topological relations was put forth, and the algorithm SCGML-LRI for spatial clustering in GML data based on line-region inclusion relations was proposed. This algorithm considered the inclusion relations between line and region spatial objects as the similarity measurement criteria. The CLOPE algo- rithm was used for clustering of spatial objects. The experimental results showed that algorithm SCGML-LRI was effective and efficient.
出处
《山东大学学报（工学版）》
CAS
北大核心
2009年第2期21-25,共5页
Journal of Shandong University（Engineering Science）
基金
国家自然科学基金资助项目(40771163
40871176)
关键词
空间聚类
拓扑关系
线面空间包含
GML
spatial clustering
topological relation
line-region spatial inclusion
GML
