Nat'l Research Council report on data mining report: it doesn't work

[David Farber <dave () farber net>]

http://news.cnet.com/8301-13578_3-10059987-38.html?part=rss&subj=news&tag=2547-1_3-0-20


October 7, 2008 9:30 AM PDT
Government report: Data mining doesn't work well
Posted by Declan McCullagh

The most extensive government report to date on whether terrorists can  
be identified through data mining has yielded an important conclusion:  
It doesn't really work.

A National Research Council report, years in the making and scheduled  
to be released Tuesday, concludes that automated identification of  
terrorists through data mining or any other mechanism "is neither  
feasible as an objective nor desirable as a goal of technology  
development efforts." Inevitable false positives will result in  
"ordinary, law-abiding citizens and businesses" being incorrectly  
flagged as suspects.

The whopping 352-page report, called "Protecting Individual Privacy in  
the Struggle Against Terrorists," amounts to at least a partial  
repudiation of the Defense Department's controversial data-mining  
program called Total Information Awareness, which was limited by  
Congress in 2003.

But the ambition of the report's authors is far broader than just  
revisiting the problems of the TIA program and its successors.  
Instead, they aim to produce a scholarly evaluation of the current  
technologies that exist for data mining, their effectiveness, and how  
government agencies should use them to limit false positives--of the  
sort that can result in situations like heavily-armed SWAT teams  
raiding someone's home and shooting their dogs based on the false  
belief that they were part of a drug ring.

The report was written by a committee whose members include William  
Perry, a professor at Stanford University; Charles Vest, the former  
president of MIT; W. Earl Boebert, a retired senior scientist at  
Sandia National Laboratories; Cynthia Dwork of Microsoft Research; R.  
Gil Kerlikowske, Seattle's police chief; and Daryl Pregibon, a  
research scientist at Google.
<snip>


-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com