go to HorizonZero HorizonZero 06 vertical line layout graphic franšais >  

printer friendly version of article  >

design : thinking with vision
View this article in flash  requires flash 6 >

Thinking with Vision
Tamara Munzner wants to make the world a better place with Infovis
by Angus Leech with Kurtis Lesick

Math visualization is fun. At least it seems to be when Tamara Munzner is in charge. Unbelievers should go to the UBC computer scientist’s Web site and download a free copy of KaleidoTile, a software tool she helped create while she was Technical Staff at the University of Minnesota’s Geometry Centre. Although ostensibly developed to explore three-dimensional symmetry in non-Euclidean and spherical geometry, the verbal reaction most people have when viewing the software for the first time is more often “Beautiful!” or “Trippy!!!” than, “Wow, I’ve never perceived the relationship between Platonic and Archimidean solids quite so clearly before!” Nevertheless, even if people do commonly see the aesthetics of Munzner’s visualizations before grasping the principles she’s trying to communicate, there’s nodoubt in her mind about what she intends to deliver to her audience — or about distinctions between the science of visualization and the practice of art.

“The goal is not to entertain but to explain. But obviously if you bore them to pieces, it’s not explaining anything. So it’s not that we don’t care about issues of timing and pacing. Also, if it’s so ugly that they cant stand it, then they won’t learn either: issues of aesthetics, colour choice and composition are all important. But our goal is explanation…My take on it is: There is an entire body of knowledge in art that has accumulated over thousands of years, and it’s a great resource for us in visualization to mine…but only some of the information is relevant. Ignoring the history of aesthetics would be moronic, but we have different goals than artists. We don’t generally evaluate our work along the same lines.”

KaleidoTile is a home-computer adaptation of Triangle Tiling, a project Munzner originally developed as a public exhibition for the Science Museum of Minnesota. And while many of her visualization projects have been intended as tools to assist research scientists or mathematicians in doing their work, many others have been created for broader educational consumption. Another perfect example is Outside In, a (frequently hilarious) Geometry Centre video aimed at bringing the wonders of contemporary mathematics to schools and the general public via the illustrative magic of multimedia animation.

Outside In is a video about trying to turn a sphere inside out without poking a hole in it, or pinching or creasing it,” explains Munzner. “It’s something that’s pretty hard to explain without moving pictures. There was an article about it in Scientific American in the Sixties that had meticulous drawings, but it was still very hard to sort out, because still figures just couldn’t communicate this process very well.”

After working at the Geometry Centre for five years on various visualization software systems (eg., Geomview) and videos (The Shape of Space; Four Polytopes and a Funeral), Munzner left in 1995 to pursue a PhD in Computer Science at Stanford University. There, her work focused on adapting computer graphics for the visualization and interactive navigation of large graphs and networks. She started wrestling with questions about how to graphically represent huge bodies of data — for example, the hyperlink structures of large Web sites, or North American Telcom databases (where dialing pattern data is kept for a year, and calls average 500 million per day) — in meaningful ways. She realized that sometimes interactive data visualization is the only hope for understanding such masses of information without disorientation. Despite its fledgling status as a discipline (or perhaps because of it), the challenges of the field excited her. And that is how, in her own words, she “fell in love with infovis”.

What exactly does Munzner mean by “infovis”? In her profile for the University of British Columbia, where she recently landed a job as Assistant Professor, she frames it as a very recent reaction to the fact that the amount of data we have access to is expanding beyond the readily comprehensible. In fact, it’s “exploding even faster than the growth rate of computational power. The field of computer-based information visualization, or infovis, is about creating tools that exploit the human visual system to help people explore or explain data. Interacting with a carefully designed visual representation of data can help us form mental models that let us perform specific tasks more effectively.”

Infovis is a new field that gleans its theoretical foundation from areas as diverse as computer graphics, human-computer interaction, cognitive psychology, semiotics, graphic design, cartography and visual art. But Munzner is the first to admit that not everyone agrees about the discipline’s definition. As guest editor for a recent IEEE Computer Graphics and Applications Special Issue on Information Visualization (January/February, 2002), Munzner explained that there is often confusion between infovis and the somewhat older area of scientific visualization.

“Information visualization hinges on finding a spatial mapping of data that is not inherently spatial,” she wrote in her editorial, “whereas scientific visualization uses a spatial layout that’s implicit in the data.” So, scientific visualization is about finding ways to augment our ability to see data that already exists in three dimensions — like “things that are on timescales too fast or slow for the eye to perceive, or structures much smaller or larger than human scale, or phenomena such as x-rays or infrared radiation that we can’t directly sense.” Time lapse representations, electron microscope images, and radio astronomy space-scapes would all therefore be examples of scientific visualization.

Infovis, on the other hand, is about representing the abstract. As Munzner’s explains, “a typical information visualization data set would be a database of film information, with the title, length, year of production, and genre for each film.” Another example would be Munzner’s recent work with evolutionary biologists trying to visualize gargantuan evolutionary trees (a project that also has potential educational applications: for example, helping schoolkids understand how Cretaceous dinosaurs relate biologically to their pet budgie or dog). Her attempts to visualize tree-like branching data sets have already spawned software systems including HypViewer (a 3D visualization tool downloadable from her Web site) and the soon-to-be-released AccordionTrees —systems which may be terribly useful indeed, but are also (it has to be said outright) a lot of fun to play with.

This capacity to entertain is fortunate, especially if Munzner is accurate in her prediction that infovis systems will become steadily more pervasive as the data we’re dosed with continues to explode. And she’ll no doubt be having vast amounts of fun herself in her new professorial role at UBC. She notes that she’s excited about having joined a Canadian university where she’ll have more opportunity to do hands-on work than is typical for academics in similar positions to the south (where larger departments mean more management of graduate students, and less personal research). In Vancouver, she’ll have plenty of time to pursue data visualization as a means for helping people navigate wide sargasso seas of information. Or, as she says with a wry smile in person, to “change the world through infovis”.

Angus Leech is the English Editor of HorizonZero. He met Tamara Munzner last summer at a conference on visualization held at The Banff Centre. Kurtis Lesick was also there, and gathered the interview that forms the backbone of this article.

back to top back to top  


Valid XHTML 1.0!
Valid CSS!