GAO: DHS does more data mining than it reports

How much data mining does the Homeland Security Department do? DHS says a little; the Government Accountability Office says more than a little, in a report dated Sept. 7 that wasn't posted online until Oct. 7.

Under the Federal Agency Data Mining Reporting Act of 2007, agencies must annually report to Congress their usage of pattern-based searches of their databases. That is, when agencies mine for data elements that depart from a predetermined set of associated behaviors, or "patterns might be regarded as small signals in a large ocean of noise," as a 2008 National Research Council report put it.

"Subject-based queries are explicitly excluded," from the requirements of the act, says the DHS official response to the GAO report, which was signed by Jim Crumpacker, head of the DHS GAO/OIG liaison office.

Subject-based queries search databases for information using a specific identifier, such as social security number or license plate.

"In practice, many data-mining systems use a combination of pattern-based and subject-based queries," the GAO report says.

Read more.

Read GAO Report: Data Mining: DHS Needs to Improve Executive Oversight of Systems Supporting Counterterrorism.