From the previous chapter, it became clear that global ranges are not always suitable for applying a rulebase in spatial applications and that better ranges can be determined. Here, a modification to allow a fuzzy rulebase system to be used for processing spatial data is presented. This is a modification to the algorithm for construction a rulebase from examples, eliminating the requirement that the domains of the variable need to be known in advance, provided it can be calculated for every variable. This also impacts the application of the rulebase, where the domains are also not a priori known. Unlike with a normal rescaling, the resulting fuzzy set is correctly defined on the domain of the result, which is beneficial for subsequent processing.