Spatial Clustering for Mining Knowledge in Support of Generalization Processes in GIS

摘要: This paper proposes spatial clustering to be used for mining knowledge in support of generalization processes GIS. Three aspects are investigated the study. First, facilitates detection hidden structures and patterns (with source map or database) that should retained course generalization, particular when multiple attributes from geometric, topological semantic perspectives involved. Second, hierarchical supported by a dendrogram provides an efficient effective tool visual interrogation exploration clusters multirepresentation. Third, some objective criteria quality assessment terms how derived represent initial dataset. All these illustrated with conducted case studies this paper.