Text by Paul Tumarkin, Tech Launch Arizona
UA Department of Computer Science
UA faculty member Stephen Kobourov has earned the Humboldt Research Fellowship, which will fund an 18-month research stint in Germany.
Stephen Kobourov, who specializes in taking large amounts of complex information and translating it into meaningful maps and graphs, has earned a Humboldt Research Fellowship to advance his work.
Kobourov, a University of Arizona associate professor of computer science, will spend 18 months in Germany on the research grant, which amounts to about $100,000.
"Presentation can make a big difference," Kobourov said. "Knowledge discovery is a very popular topic and with so much data, can we discover something new in this data?"
Kobourov is among 587 fellowship recipients worldwide, including 97 from the United States. All are currently faculty members or postdoctoral students working across a range of disciplines, Germany's Alexander von Humboldt Foundation reports.
Fellowship recipients are selected on the basis of their academic and scholarly work and are considered "highly-qualified scientists and scholars," the foundation also noted.
Kobourov will be on leave from the UA from August 2011 through July 2012, hosted by the University of Tübingen in Germany. There, he will spend his time advancing data visualization through his project, "Putting Data on the Map," while also giving talks at other institutions.
A relatively new field in computer science, data visualization involves taking highly complex bits of scientific information and translating it into dynamic visuals, whether in the form of graphs, maps or animations.
To better understand the process, consider your social network. Conventional models tend to be one-dimensional, merely grouping things and draw connections between points – like linking Francis to Tina to Thomas to Luis to Gabrielle.
"It's an object and a relationship. That's what I used to do, until recently," Kobourov said.
Today, his work is more involved, bringing clarity not only to linkages, but other connections. For example, based on Kobourov's mapping of Amazon.com, it appears that Emily Brontë's "Wuthering Heights" may serve as one of the gateway books for readers who end up developing a following for vampire-themed novels.
Likewise, in the land of music, people who enjoy the likes of Beyonce and Janet Jackson may be connected to musical types like Thievery Corporation and Massive Attack through the sounds of Sia and Emiliana Torrini.
"The only thing I've changed is the grouping," Kobourov said. "When you present information like this, people start to see stuff that was there before, but that they didn't see before."
While in Germany Kobourov intends to expand such maps, focusing more keenly on the Netflix movie database, DNA comparisons and other datasets he finds through connections he makes abroad. He also eventually plans to map the Internet and the blogosphere.
The implications for such work vary, helping not only to better understand information, but to develop better questions around different types of data.
"There is a standard cliché that says a picture is worth 1,000 words," Kobourov said. "We have evolved to be visual. Often, you can see things faster and draft an idea quicker if you are looking at the similarities between points."
Implications also exists for scientists and for companies, he added.
"There are companies monitoring what is being tweeted about them so that they can respond quickly before something goes into the mainstream," Kobourov said. "What better way to show it than with a nice map?"
Kobourov has earned numerous other awards and grants for his research contributions, including the National Science Foundation Career Award and a Fulbright Scholarship.
Earlier this year, he and his collaborators earned an 8-month, $380,000 grant from the U.S. Office of Naval Research for "Putting Network Security on the Map: Visualizing Network Security with a Unified Map Metaphor."
Other UA researchers on the project are: Christian Collberg, associate professor of computer science; Loukas Lazos, an assistant professor of electrical and computer engineering; and Srini Ramasubramanian, an associate professor of electrical and computer engineering.
Kobourov is also part of another team on a U.S. Department of Agriculture grant totaling more than $910,000. Team members on the project, "A Collaborative Mind-Mapping Solution to the Obesity Challenge," include Randy Burd, an assistant professor of nutritional sciences, and Melanie Hingle, an assistant research scientist of nutritional sciences.
Above all, Kobourov wants to continue to present information in more compelling ways while developing methods and algorithms that can help people better understand data.
"Underneath, there is all this technical, scary stuff and geometric problems," he said. "What I am trying to do is to solve problems that are kind of practical and, in the process, also solve these more technical and theoretical problems."
UA Department of Computer Science