Attensity finds scattered data
- By Brian Robinson
- Oct 12, 2004
A company partly funded by the CIA has introduced a new version of an application suite aimed at extracting useful information from unstructured data, a rich source of information that's so far been left mostly untouched by data-mining tools.
Although most extraction tools focus on the data in structured environments such as databases, experts say up to 80 percent of the useful information in any organization may be exist in e-mails, field reports, service records, news feeds and other disparate documents. Until now, there's been no way to get useful data from these sources, said Craig Norris, chief executive officer of Attensity Corp., a 3-year-old firm whose investors include In-Q-Tel Inc., a nonprofit company funded by the CIA to identify and invest in cutting-edge technologies.
"Our technology allows users to go through unstructured text and get the who, why, what and where of the information that's included," he said.
"Attensity's sentence diagramming and parsing techniques allow them to get into the key relationships enclosed in the text."
That information can be fused with structured data to give a much deeper understanding of the whole information collected by an organization.
The Attensity Application Suite includes new versions of Attensity's core technology and a line of business-specific applications, including ones for government intelligence, law enforcement and logistics. The suite includes the ability for users to automatically share data and insights from a central data repository.
"That was the No. 1 [request] the intelligence community in particular came back to us with," Norris said.
In-Q-Tel officials chose Attensity as one of its investments because there was a need for a way to look at unstructured data, and no other company was really focused on that, said Greg Pepus, director of In-Q-Tel's federal and intelligence community strategy.
"Attensity is good at understanding the context of what's a noun, pronoun etc.," he said. "So it can reliably dissect a document and then put the relevant elements into a table for you."
Attensity is only one piece of the unstructured data puzzle, Pepus said. "But we feel it's a major step forward."
Attensity has also been added to the Defense Department's Intelligence Information Systems Integration and Engineering Support Services Contract 3 as one of the teams that prime integrator Booz Allen Hamilton Inc. oversees. A formal announcement of that is due later this month.
Brian Robinson is a freelance journalist based in Portland, Ore. He can be reached at [email protected]
Brian Robinson is a freelance writer based in Portland, Ore.