Abstract (EN)

Work on multiscale issues presents countless challenges that have been long attacked by GIScience researchers. Research is usually concentrated in one of two directions - new data models to support handling multiple scales, or data structures and algorithms to process data across scales. Complementary implementation aspects are concerned with generalization (and/or virtualization of distinct scales), or with linking entities of interest across scales (e.g., using bottom-up implementation of specific structures, without relying on any specific DBMS). However, researchers seldom take into account the fact that multiscale scenarios are increasingly constructed cooperatively, and require distinct perspectives of the world, in which each research group considers specific aspects of a problem. The combination of handling multiple scales at a time, and having multiple user perspectives per scale constitutes what we call multi-focus research. This paper presents our proposal to attack multi-focus scenarios, which considers distinct aspects of the problem of managing multiple scales, illustrated with examples of multiscale geospatial data. Our approach builds upon a specific database version model -- the so-called multiversion MVDB -- which has already been successfully implemented in several geospatial scenarios, being extended here to support multi-focus research. This extension was implemented and tested in a real world case study, briefly discussed here.