Reasoning with words about geographic information

Geographic information systems have gained an increasing interest over the recent years. However, their abilities are restricted in that they usually reason about precise quantitative information only, which means that they fail whenever exact matches cannot be found. They do not allow for any form of reasoning with imprecision. In this article, we describe a way of incorporating imprecise qualitative spatial reasoning with quantitative reasoning in geographic information systems. In particular, we show how tessellation data models can be extended to allow for qualitative spatial reasoning. The idea is to associate qualitative information with fuzzy sets whose membership grades are computed by applying the concept of proximity. In addition, we will show how images like geographic maps or satellite images can be analyzed by computing the distances between given reference colors and the colors that occur in the image, and how the results of this analysis can be used in the fuzzy spatial reasoner.