Danyel Fisher, Steven Drucker, Roland Fernandez, and Xiaoji Chen
Finding ways for information workers to easily create and modify visualizations that display their own data has been a long time goal within the visualization community. We describe Vis-à-vis, a declarative language for defining and extending visu-alizations directly within spreadsheets. Vis-a-vis allows users to directly bind data and formula to the visual attributes of an extensible set of visualization primitives. The visualizations that Vis-a-vis creates can be shared and modified easily, allowing users to modify existing visualizations. This approach allows users to select visualizations from a gallery, to customize them easily, or to create novel visualizations. The approach leverages familiar formulas and data from spreadsheets. We prototype a system that uses this language, and use it to build a number of standard and custom visualizations, and gather formative feedback from a small user study.
Publisher Microsoft Technical Report