Visual analysis of spatiotemporal multilevel data

Abstract — For multilevel data, levels are not the result of a hierarchical aggregation, but contain independently produced data. While first visualization techniques for this kind of data exist, suitable interactive exploration techniques have rarely been investigated so far. In this paper, we introduce means for the analysis, representation, and exploration of spatiotemporal multilevel data as first solutions to fill this gap. Particularly, these are analysis strategies serving to order the data appropriately, a novel data display providing an overview to the data as well as level- and path-based exploration strategies supporting systematic data browsing. Appropriate level and path filtering techniques are also provided. Results from an application of this technology to election data indicate their usefulness for conducting visual analysis tasks in spatiotemporal multilevel data.