Towards a Flexible Framework for Regionalization

A theoretical framework is proposed to analyze regionalization problems consisting of quantitative and qualitative information with varying degrees of imprecision. Fuzzy set concepts of distance and similarity are formulated to construct regional grouping procedures. Depending on the data and the concepts employed, regions derived may be mutually exclusive or they may overlap to a certain degree. Hierarchical systems of regions are also analyzed on the basis of asymmetric spatial relations. Compared with conventional methods, this framework appears to be more natural and flexible in preserving information pertinent to regionalization.