One man's spam is another's art

For the last several years, the Romanian-born computer artist has applied techniques in computational modeling and information visualization to invent a new form of artistic expression. One of his more notable projects involved creating what he calls Spam Plants. He wrote algorithms that analyzed various text and data points of junk e-mail to produce "organic" images of plantlike structures that spontaneously grew based on incoming spam.

Now he's working on a software agent that can "write" experimental graphical novels based on a melange of text culled from thousands of like-minded blogs across the Net. When finished, the agent, called Blogbot, will include algorithms that apply computational linguistics to the blog text, so that it extracts meaning from the text. That way, the graphical novel might strike on profundity.

"By analyzing text using computational linguistics methods, you can detect anger and sadness. Turning those into gestures in three dimensions, that would be interesting," said Dragulescu, who is now head of the Experimental Game Lab, a research lab at the Center For Research and Computing and the Arts (CRCA) within the University of California at San Diego.

Dragulescu's work stands out at a time when scientists and technologists are struggling to harness the massive quantities of data online and make sense of it for fields like earth science, drug discovery, genetic research and U.S. security. Information visualization is a traditional scientific method that's getting a lot of attention now because it involves projecting a visual image of the data so that onlookers can make connections that might otherwise be lost.

Making such conceptual leaps is often associated with art, and for Dragulescu, that's the point. Using scientific visualization methods to make art shows how technology changes art, he said.

"My efforts (have been) to expose the ubiquitous forms in which data and technology are both actively and passively shaping the ways we perceive and construct ourselves and others," Dragulescu said.

Spam, for example, has so blemished and clogged an otherwise revolutionary new form of communication that scientists have taken to storing and studying it in recent years. That's why, Dragulescu said, he decided to analyze it for artistic purposes and try to expose the hidden nature of junk e-mail, with its duplicitous headers and subject lines.

He doesn't use Photoshop but simply writes code to create computer art. For the Spam Plants, he parsed the data within junk e-mail--including subject lines, headers and footers--to detect relationships between that data. Then he visually represents those relationships.

For example, the program draws on the numeric address of an e-mail sender and matches those numbers to a color chart, from 0 to 225. It needs three numbers to define a color, such as teal, so the program breaks down the IP address to three numbers so it can determine the color of the plant. The time a message is sent also plays a role. If it's sent in the early morning, the plant is smaller, or the time might stunt the plant's ability to grow, Dragulescu said.

The size of the message might determine how bushy the plant is. Certain keywords, such as "Nigerian," might trigger more branches. But Dragulescu did not inject any irony. Messages about Viagra do not grow taller, for example.

"I mapped it in a way to control the complexity of the plant," he said.

History lesson
Dragulescu moved from Romania to the United States in 1997 to study film, photography and art history at Ithaca College. While there, he taught himself computer animation and programming, and after graduation, took a job in San Francisco writing code and working on Web, multimedia and game design. That's when he seized on the idea of using algorithms to harness the Net's mountains of data and produce art.

After the dot-com bust, Dragulescu earned a master of fine arts degree from the University of California at San Diego. There, he studied computer science, cognitive science, media theory and visual arts and was inspired to experiment with art and data from the Net, including spam, music and blogs.

One project from CRCA is called Scalable City, which creates procedural game assets and environments.

Now he's working on taking the same principles and applying them to text and literature, so that one day he can create a visual image of a Hemingway novel, for example.

He recently completed a project in music visualization called Extrusions in C Major. The project creates images from Mozart compositions. To do this, the software analyzes the note characteristics of the music, including the tempo of various instruments, and then pairs them with colors: white for piano, yellow for violin and blue for cello.

With Blogbot, though it's not finished, he's already created a first graphical novel, called "What I Did Last Summer." The piece draws on writings about the war in Iraq from soldiers who have maintained blogs. To make the program even more sophisticated for his next novel, Dragulescu is learning computational linguistics methods to understand the logic and emotions embedded in language.

Strangely enough, his interest in language analysis started with spam.

"Spam is a random piece of literature, it has unseen effects, it changes all the time," he said. "And it's led me to see text differently."