Spatial Similarities in Urbanisation and Regional Diversity

Abstract

Most of the large databases currently available have a strong geospatial component and contain potentially useful information that might be of value. Data mining is defined as the inspection of data with the aim of discovering knowledge. Mining implies a laborious process of searching for hidden information in a large amount of data. Knowledge discovery is defined as the discovery and formal representation of knowledge from data collections. Data mining in connection with knowledge discovery techniques will be of increasing importance for the urban research and planning processes. »Urban Data Mining« is a methodological approach to discover logical, mathematical and partly complex descriptions of patterns and regularities inside a set of data. The main theme of this contribution is the definition of an urbanized area. 12430 German communes are structured by a classification approach. The issue of these grouping processes are urbanisation and regional diversity. The set of data is examined and it will be shown that the investigation of distributions leads to a better understanding of each attribute. Gaussian Mixture Models are presented as an appropriate method for regionalisation. Spatial Analysis (GIS) is part of the process of knowledge conversion and communication. Results suggest a general typology and can lead to the development of prediction models using subgroups instead of the total population. The procedures on the basis of knowledge-based systems are currently not sufficiently developed for a direct integration into the regional and urban planning and development processes. These approaches could lead to a benchmark system for regional policy or to other strategic instruments such as fully automated urban monitoring systems.