[Total: 0    Average: 0/5]

The GIS revolution and the increasing availability of GIS databases emphasize theneed to better understand the typically large amounts of spatial data. Clustering isa fundamental task in Spatial Data Mining and many contributions from researchersin the field of Knowledge Discovery are proposing solutions for class identificationin spatial databases. The term spatial data refers to a collection of (similar) spatialobjects, e.g. areas, lines or points. In addition to geographic information, eachobject also possesses non-spatial attributes. In order to apply traditional data miningalgorithms to such data, the spatial structure ans relational properties must be madeexplicite. SCOT deals with the special case of grouping German towns. The townsare related to each other by the various streets connecting them. Each town alsopossesses an inner spatial structure, the local street network, and further non-spatialinformation. This thesis considers all three kinds of information for the clusteringof towns. It exploits the concept of neighborhood to capture relational constraints,measures the similarity of the structures of local street networks and transforms themost important non-spatial attributes. SCOT is part of a project at Fraunhofer IAIS,Germany, and has been successfully applied in practice.