Authors: Katarína Furmanová, Miroslava Jarešová, Bikram Kawan, Holger Stitz, Martin Ennemoser, Samuel Gratzl, Alexander Lex, Marc Streit
Abstract: Tabular data visualizations are easy to understand and powerful at communicating patterns in datasets, especially when paired with interactions such as sorting. In this work we present Taggle, a novel visualization technique for large and complex tables. We consider datasets that are composed of columns of categorical, or numerical data, in addition to homogeneous matrices. The key contribution of Taggle is its ability to aggregate data subsets (rows and columns) on demand based on user-defined grouping rules. Different visual representations for individual cells and aggregated subsets are available. The aggregation strategy is complemented by the ability to sort hierarchically, and by rich data selection and filtering capabilities. We demonstrate the usefulness of Taggle using an AIDS dataset for 160 countries.