EXTENDING THE INDIVIDUAL FORMALISM FOR A MORE COMPLETE MODELING OF URBAN SPATIALLY REFERENCED DATA

The modeling formalisms currently used to design Spatial Information S ystems (SIS) databases provide designers with limited possibilities in the modeling and processing of data. Besides being inefficient in des cribing spatial reference, these formalisms neglect temporal reference (existence and evolution of objects), give rise to complex models, an d isolate processes from the data. In the case of large urban organisa tions, these limitations create technical difficulties when SIS are be ing designed. Designing can be facilitated by improving the expressive richness of these models. The goal of this paper is to present an ext ension to the Individual Formalism (used in the MERISE method) which p ermits a better description of spatially referenced databases. After h aving summarized the current problems with existing formalisms, we wil l present the different solutions elaborated and integrated into a new formalism called MODUL-R. Certain elements of MODUL-R which are parti cularly useful in urban environment applications are emphasized. These elements will be illustrated with examples taken from an urban SIS pr ototype, to be followed by a consideration of the potential of MODUL-R .