Data mining humming along at DHS

Although the Homeland Security Department terminated a controversial visual analytics data mining program this summer, it continues to engage in visual analytics research in a separate program, a spokeswoman confirmed.

The ongoing visual analytics research at the Science and Technology Directorate is being publicized as a means of eventually identifying terrorists through potential use of data collected from video surveillance footage, cell phone calls, photos, bank records, chat rooms and e-mails. But no real-world, operational data is actually being used in the research, said DHS spokeswoman Amy Kudwa.

"It relies on synthetic data,' Kudwa said. "It is purely research on ways to interact with data."

Visual analytics is considered a form of data mining, which is defined as use of computer programs to find hidden patterns in large amounts of data and to use those patterns to predict behavior. Data mining is widely used in commerce, but it has been controversial in homeland security because of the fears of privacy loss and civil-rights violations. In visual analytics, the data is mapped in two- and three-dimensional formats and sometimes animated.

Science and Technology's visual analytics research program uses only fake data rather than real-world, operational data, and as such is not considered data mining, according to Kudwa. "There is no data mining aspect."

The research is separate from the directorate's former Analysis, Dissemination, Visualization and Semantic Enhancement, or Advise, data mining program, Kudwa said. That program, which also used visual analytics, was shut down this year. The DHS Privacy Office and the inspector general's office have criticized the Advise program for not fulfilling privacy requirements. It was terminated because it completed its research cycle, Kudwa said.

Although the ongoing research program uses synthetic data, the Science and Technology directorate, in a recent newsletter, described it as having a potential for identifying and stopping terrorists.

"Today, researchers at the DHS Science and Technology Directorate are creating ways to see fuzzy data as a three-dimensional picture where threat clues can jump out," a recent DHS newsletter stated in an article on the visual analytics program.

"Mathematicians, logicians and linguists make the collective universe of data assume a meaningful shape. They assign brightness, color, texture and size to billions of known and apparent facts, and they create rules to integrate these values so threats stand out. For example, a day's cache of video, cell phone calls, photos, bank records, chat rooms and intercepted e-mails may take shape as a blue-gray cloud," the DHS newsletter said.

The newsletter also suggests the program may be used eventually to predict behavior. "A month of static views might be animated as a 'temporal' movie, where a swelling ridge reveals a growing threat. Analysts can then state, 'I think a bomb will explode here,'" the newsletter stated.

The visual analytics research has been ongoing since 2004, Kudwa said. However, it was not listed as one of 12 data mining programs in the department in an August 2006 survey by DHS Inspector General Richard L. Skinner.

The ongoing visual analytics research is headed by Joseph Kielman, basic research lead for the directorate's Command, Control and Interoperability Division. Kielman advises the National Visualization and Analytics Center based at Pacific Northwest National Laboratory.