Inspired by Gene Davis’s style and using World Bank data from 2010, this piece portrays the urban population and annual growth for 36 countries (vertical bands) with populations of 20 million or more. Cell width indicate urban population count. In pink is Ukraine, the only country with negative urban population growth. Other countries are colored in a black to blue scale, where black represents zero urban population growth and blue represents the highest among these countries (6.25%).

The Evocative Charm of Treemaps

The first time I saw Ben Shneiderman’s Treemap Art Project pieces, which he has been producing over the years, was during a visit to the University of Maryland in April 2014, when I gave a lecture on Visualization Metaphors. As Ben walked me through many of the pieces in person, it was clear how spirited he became when explaining every seemingly subtle choice, from the most suiting layout orientation, to the most evocative color scheme.

Even though this expressive venture might seem odd for a renowned computer science scholar, considered the father of the modern, computer-generated treemap, Ben Shneiderman has always been cognizant of the treemap’s pictorial force and was bold enough to pursue it. As he explains:

Although I conceived treemaps for purely functional purposes (understanding the allocation of space on a hard drive), I was always aware that there were aesthetic choices in making appealing treemaps, such as design, color, aspect ratio, and the prominence of borders for each region, each hierarchy level, and the surrounding box. In addition, certain treemaps are inherently interesting because of the data displayed or patterns revealed.

Inspired by Piet Mondrian’s compositions this piece depicts the popularity of the top 20 musicians, based on user data from Last.fm. The size of a cell represents the number of times their tracks were played, while color indicates genre: rock (white), alternative (blue), pop (yellow), hip-hop (red).

More recently, 0n October 16th, 2014, I had the privilege of being part of the DASER discussion panel at the National Academy of Sciences (Washington D.C.) with JD Talasek, Jonah Brucker-Cohen, Jon Froehlich, and Ben Shneiderman. This event was organized in conjunction with the opening of Ben Shneiderman’s Every AlgoRiThm has ART in it: Treemap Art Project Exhibition, a showcase of his 12 enticing prints that will be on display at the Keck Center until April 15, 2015.

This piece depicts statistics about a selected group of TED talks. The dataset was compiled by Sebastian Wernicke for his TED talk on “Lies, Damned Lies and Statistics.” The size of each rectangle represents the engagement score for a certain TED talk, while colors convey the total number of del.icio.us bookmarks. In an effort to capture the variety of TED talks, the coloring scheme goes from red (lowest) to pink (highest).

I also had the pleasure of writing a short essay for the exhibit’s accompanying booklet, which you can read below.

Essay

History is filled with symbiotic exchanges between art and science, and more specifically between art and cartography. David Woodward, in his Art & Cartography, published in 1987, describes in detail how numerous artists were influenced by cartography, and how maps themselves, as witnessed in the work of numerous Dutch painters, were hung on walls as pieces of art.

Ben Shneiderman’s eminent work in creating a tiling mechanism to visualize a tree structure by means of nested rectangles was a pivotal contribution to the infant domain of data visualization in the early 1990s. The resourceful model devised by Shneiderman, and later identified as a treemap, quickly became one of the most widespread methods for visualizing hierarchy, primarily due to its efficient and adaptable use of space. Perhaps most importantly, the pioneering nesting algorithm opened the door to a huge array of new visualization models and approaches.

The original rectangular treemap, like many of its subsequent variations, such as the Voronoi treemap and the circular treemap, is not just effective at mapping tree structures. All models without exception share a remarkably expressive quality. Their elaborate cell layout and colorful display, which frequently looks like intricate stained glass, is as visually captivating as it is pragmatically functional. Treemaps not only epitomize the recent growth of data visualization but also its occasional, enthralling aesthetical charm.

Shneiderman’s exhibit “Every AlgoRiThm has ART in it: The Treemap Art Project” gathers twelve compelling pieces of evidence for the treemap’s pictorial force. Some of the pieces intentionally pursue the lure of other artists’ style, such as The Singing Mondrian, while others like Dazzling Talks have their own unique, magnetic charisma. The combination of colors, the scaling and subdivision of cells, the orthogonality and confluence of its multiple lines, create an enticing mesh that resembles at times the essentials of form and color found in neoplasticism, orphism, minimalism, and abstract expressionism. All pieces can immediately draw you in, as you marvel at the details of its intricate matrix. Once you realize each piece is depicting real data attributes from a specific subject, like international flight routes or the individual scoring of basketball players, then its allure becomes stronger, more powerful, and surprisingly more meaningful.

With this influential exhibit, Ben Shneiderman completes the full circle of the treemap’s adaptability, first as a revolutionary utilitarian tool to map large hierarchies, and now as a powerful visual element able to entice and evoke emotions. Most importantly, the exhibit successfully bridges the gap between art and science, by showing they do not represent two sides of the same coin, but one single side, or should we say algorithm, seen through different lenses.

Inspired by Hans Hofmann’s The Gate (left), Ben Shneiderman’s Frequent Flyers (right) showcases the ratio of international to domestic flights for individual airports. Larger cells reveal more international flights, while colors represent the variation in total number of routes served by that airport — green for less busy airports. The dataset was collected from openflights.org.