An application-oriented view on graded spatial relations

Approaches for modeling graded spatial relations abound in the literature on image processing and on geographic information systems. In contrast, few proposals have addressed the use of degrees in application contexts where inference plays a central role. We argue that the use of degrees is nonetheless natural in such contexts, and may enhance the application potential of qualitative spatial reasoning in domains where relations between vague regions need to be expressed, as well as domains where the robustness of spatial relations w.r.t. small changes in the underlying configurations is important. In particular, we discuss the interpretation of degrees, contrasting fuzzy regions with formal accounts of spatial vagueness based on supervaluation semantics. We also touch upon the problem of acquiring (fuzzy) spatial relations from the web in an automated way.