TreeVersity gives users a powerful set of tools to detect node values changes and topological differences in tree comparisons. The design employs complex color palettes to show positive/negative, absolute, and relative change, shapes that pre-attentively showed these changes, and novel graphical metaphors that highlight changes. Collaborators: John A. Guerra Gómez (PI), Ben Shneiderman, Catherine Plaisant of the UMD Human-Computer Interaction Lab (HCIL) and Michael L. Pack with the UMD Center for Advanced Transportation Technology (CATT) Laboratory.

Visit Site →

TreeVersity is a collaboration comparison tool for analyzing big data. It allows users to detect node values changes and topological differences. I served as a design consultant for this collaboration with Dr. John A. Guerra Gómez (PI) and UMD Human Computer Interaction Lab.

View of data visualization

TreeVersity’s main interface uses a novel interactive information visualization technique to display the differences in the tree, combined with coordinated complementary views, and controls to filter the data and identify insights.

Another view of the data visualization

The tool allows access to multiple views of the data as with the scatter plots in the lower right and the overall comparison between two data sets and then detailed information about those changes in the visualization below. Here we are showing all the Agencies, Bureaus and Accounts that change in more than $12 Billion dollars in the U.S. Federal Budget between 2011 and 2012.

Another view of the data visualization

After identifying interesting sub-trees in the Overview mode, users can navigate into them to explore local changes. The image on the top shows the change in number of passengers in the United States by month and State between 2001 and 2000, sorted by absolute change. The middle figure shows how users can explore different aspects of that data, as with the number of passenger decreases, and then look at more detailed analysis, as with the comparisons of airport travel after 9/11.

Another view of the data visualization

The Bullet representation: DiffTree construction. For each node, a “bullet” is created in the DiffTree whose size represents the absolute change and the color the relative change on it.

Another view of the data visualization

Filtering by change: TreeVersity also allows users to filter nodes by the amount of relative or absolute change using intuitive controls. In this example, the user has filtered only the most stable TRB authors (those changing less than 100%, positively or negatively) by subject area. The names have been grayed out to maintain privacy.

Before view of the design

This is a version of the tool before I came on board as a design consultant.