Extending F‐Perceptory to model fuzzy objects with composite geometries for GIS

When analyzing spatial issues, geographers are often confronted with many problems with regard to the imprecision of the available information. It is necessary to develop representation and design methods which are suited to imprecise spatiotemporal data. This led to the recent proposal of the F-Perceptory approach. F-Perceptory models fuzzy primitive geometries that are appropriate in representing homogeneous regions. However, the real world often contains cases that are much more complex, describing geographic features with composite structures such as a geometry aggregation or combination. From a conceptual point of view, these cases have not yet been managed with F-Perceptory. This article proposes modeling fuzzy geographic objects with composite geometries, by extending the pictographic language of F-Perceptory and its mapping to the Unified Modeling Language (UML) necessary to manage them in object/relational databases. Until now, the most commonly used object modeling tools have not considered imprecise data. The extended F-Perceptory is implemented under a UML-based modeling tool in order to support users in fuzzy conceptual data modeling. In addition, in order to properly define the related database design, an automatic derivation process is implemented to generate the fuzzy database model.