Decision Support under Uncertainty in Spatial Contexts: The Geologic Map Creation Case

It is widely-recognized that space, together with its relation to application features, plays a vital role in several domains such as geographic information systems, multimedia systems, cartography, and remote sensing. In all these domains a multitude of rich and complex information is involved. Therefore, it is quite important to have tools for efficient representation, retrieval, and reasoning. Database systems play a central role in all aforementioned application areas, although current technology and conceptual tools do not adequately support users needs. Most of the features involved are related to subjective assessments of experimental results. As it is evident,uncertainty arises as an inherent property, which is especially important while under addressed in spatial database and related decision systems. This paper is focused on the incorporation of uncertainty in modelling and reasoning issues of such systems. The proposed approach is verified in a case study related to geology maps. 2. Approach The proposed methodology for handling uncertainty in features related to space consists of the following steps: a) Data specification and collection b) Data processing and model extraction c) Reasoning and decision support The specification and collection of data, which are application specific,result in a well-defined experimental data set. This data set consists of feature values corresponding to various locations of the area under study. Provided that the collected data set is a sample of the reality and does not describe the whole area of interest, appropriate interpolation methods are applied to assign feature values to all spatial granules.