Supporting a focus+context interaction style for spatial databases

We present results on the design of a visual interaction environment for geographic information systems. In our model a database is a spatially related collection of regions with features, hierarchically organized as a nested partition. We envision an exploration of the database where, starting from an abstract view, more detailed information on subregions and/or features is successively disclosed on demand and visualized in a fish-eye view fashion. The interaction process is a sequence of views on the database, displayed by maps preserving topological properties. Efficient navigation requires the adoption of topological invariants for representing such partitions. Since each interaction step modifies only a portion of the current invariant, the definition of our interaction primitives is based on an efficient incremental approach (the current topological invariant is updated by exploiting knowledge coming from previous steps, instead of recomputing always everything from scratch)