Visualization of tabular data---for both presentation and explorationpurposes---is a well-researched area. Although effective visual presentationsof complex tables are supported by various plotting libraries, creating suchtables is a tedious process and requires scripting skills. In contrast,interactive table visualizations that are designed for exploration purposeseither operate at the level of individual rows, where large parts of the tableare accessible only via scrolling, or provide a high-level overview that oftenlacks context-preserving drill-down capabilities. In this work we presentTaggle, a novel visualization technique for exploring and presenting large andcomplex tables that are composed of individual columns of categorical ornumerical data and homogeneous matrices. The key contribution of Taggle is thehierarchical aggregation of data subsets, for which the user can also choosesuitable visual representations.The aggregation strategy is complemented by theability to sort hierarchically such that groups of items can be flexiblydefined by combining categorical stratifications and by rich data selection andfiltering capabilities. We demonstrate the usefulness of Taggle for interactiveanalysis and presentation of complex genomics data for the purpose of drugdiscovery.
展开▼
