Use of fuzzy BK-products of relations in designing architectures of intelligent geographic information systems

The paper outlines the employment of fuzzy relational computations and fuzzy BK-products of relations in building architectures of intelligent geographic information systems (GIS) and decision support. The purpose of such systems is to analyse and interpret data to be used in geographic research. Our emphasis is on digital satellite imagery, a data stream that is commonly used to generate land use and land cover maps with which other GIS data are integrated. In order to efficiently apply methods of computational intelligence and soft computing to the project, a new conceptual framework is needed, in which the data, knowledge and activities of computing systems and human participants are jointly considered. Our framework distinguishes objects, properties, attributes, observers, contexts and perspectives and relations between these. The formal specification methods employ semiotic descriptors combined with fuzzy relational systems, algebraically defined by means of nonassociative BK-products of relation. BK-products are used not only for specifications but also in parallel computations in evaluation of data