ECAI 2004 Conference Paper

[full paper]

Qualitative Interpolation for Environmental Knowledge Representation

Antony Galton, James Hood

Environmental Knowledge Representation is concerned with representing and reasoning about the surrounding spaces, or environments, within which intelligent agents go about their businesses. A key facet of this is an understanding of qualitative features of the landscape. To understand a landscape is to have command of a structured model in which important features are highlighted and their mutual interrelations made available for inference; in constructing such a model we must acknowledge that our cognitive engagement with large-scale environments differs in important ways from the understanding of `table-top' spaces which has been prevalent in AI. We advocate an approach in which knowledge about objects (e.g., landmarks and paths) in the landscape is supplemented by knowledge of spread-out characteristics of the terrain, represented by means of spatial fluents or qualitative fields. In particular we focus on some problems of qualitative interpolation (a form of non-monotonic spatial reasoning) arising in the context of this approach.

Keywords: Spatial Reasoning, Qualitative Reasoning, Spatial Interpolation, Knowledge Representation, GIS

Citation: Antony Galton, James Hood: Qualitative Interpolation for Environmental Knowledge Representation. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.1017-1018.