Article révisé par les pairs
Résumé : Successful information management implies the ability to design accurate representations of the real world of interest, in spite of the diversity of perceptions from the applications sharing the same database. Current database management systems do not provide representation schemes that preserve each perception while fully supporting their diversity and maintaining their consistency. This is a major hindrance for building an all-embracing view of the world while serving multiple applications, whether it is by developing a single database or by providing transparent access (e.g., via the Web) to several heterogeneous data sources (that would typically hold a great diversity of stored representations). This paper reports on results from the multiple representations and multiple resolutions in geographical databases project,11In addition to the authors' institutions, the MurMur consortium included the French Mapping Agency (Institut Géographique National, Paris, France), a public research institute on Agricultural and Environmental Engineering, (Cemagref, Grenoble, France), and a GIS provider (Star Informatic, Liège, Belgium). funded by the European Commission under the 5th Framework Programme. The objective of the project has been to enhance GIS (or DBMS) by adding functionality that supports multiple coexisting representations of the same real-word phenomena (semantic flexibility), including representations of geographic data at multiple resolutions (cartographic flexibility). The new functionality enables a semantically meaningful management of multi-scale, integrated, and temporal geo-databases. © 2005 Elsevier B.V. All rights reserved.