Briefs

Technology is making it harder for the bad guys to get away.

Where police officers have used old-fashioned shoe leather, extensive interviews and sheer intuition to solve crimes, more of them also are using software developed in Tucson to scour databases for information and link key clues together.

The software is called Coplink, and it organizes information from different record management systems, ranging from arrest records to mug-shot files to even motor vehicle registrations and pawn broker records. Algorithms designed to detect patterns and relationships then help law enforcement officers rapidly identify criminal suspects, relationships between suspects or victims, and patterns of crime in an area.

The software is especially valuable in tying together information from the disparate computer systems of different agencies and jurisdictions.

The system's power relies on the belief that 80 percent of crime is committed by 20 percent of people already recorded in the criminal justice system, said Robert Griffin, president of Knowledge Computing Corp. And its return on investment is more than financial, he said.

"Is this going to take bad guys off the street? That's the ROI," he said.

The software got its start at the artificial intelligence lab at the University of Arizona. Director Hsinchun Chen was interested in using data mining and spidering technology to tie together data from different Web sites. He originally thought of biomedical applications, Griffin said, but a Tucson Police Department sergeant in Chen's graduate class spotted the idea's crime-fighting potential.

A prototype was developed in 1998 with a $1.2 million grant from the National Institute of Justice, the research-and-development arm of the U.S. Justice Department.

Knowledge Computing was created to license the technology from UA and commercialize it. The company issued its first commercial version in November, and more are in the works.

Officers from the Tucson Police Department and Coplink technicians took the system to Washington, D.C., last year to help solve the sniper shootings case. A tip from a trucker led detectives to suspects John Malvo and John Mohammed one day after the Tucson officials arrived, but Griffin says the database they built with 1 million records held eerie clues that could have found the pair, too.

A Coplink search for people and vehicles that had been stopped within 30 minutes of several shootings turned up repeated stops of Malvo, Mohammed and their blue Chevy, Griffin noted.

"It's the 80/20 rule," he said. "They're somewhere in the system."

Coplink is being used in Tucson, Boston, northeastern Kansas and other locations. By year's end, the company expects the software to be in nearly 30 locations nationwide, including multistate and regional networks.

The Phoenix Police Department will use the system and hopes to begin testing it in October, said Sgt. Randy Force, a spokesman.

The department has a rudimentary analysis system, but Coplink's offer is far above that, he said.

"It's going to be a tremendous help to investigators who are trying to locate suspects with scant information," he said.

Here is the full story from Arizona Republic.