Select the Appropriate Map Depending on Context in a Hilbert Space Model (SCOP)

Human use of categories exhibits a prototype effect; concepts become more defined through a conversation. Modelling these gradual clarifications of what a word signifies is equally important in human - computer interactions, for example in interactions about geographic concepts and the information that is needed in a given situation. We address here the simplified, but essentially realistic, question of what is meant by "map" and how the concept is refined. We apply the methods Aerts, Gabora and Rosch have described and explore how they can be integrated into practical systems. In this paper we explore the optimal selection of a map through a conversation with the client to elucidate their intentions. The example case contains effects which are similar to the "guppy effect" that is known from the literature and is a key reason to apply quantum mechanical formalism. The results are promising, and we sketch the extension to the construction of "custom made" maps from layers. This will provide users with maps that optimally reflect what map elements should be visible for use in a given context.