[EAS] Data Mining

pjk pjk at design.eng.yale.edu
Thu May 27 21:55:58 EDT 2004

Subject:   Data Mining

(from NewsScan Daily, 21 May 2004)

      Florida database contractor Seisint apparently wowed federal 
officials during a demonstration of its "high terrorism factor"
scoring  system by flagging the names of 120,000 individuals who
showed a  statistical likelihood of being terrorists. Impressed by
the scoring  technology, the Justice Department followed up by
investigating numerous  people on the list, and ultimately appointed
Seisent the sole contractor on  the $12-million Matrix (Multistate
Anti-Terrorism Information Exchange)  project. Because the Matrix
system includes extensive information on people  with no criminal
record as well as shadier characters, it has sparked  criticism from
many privacy groups and all but five states have now  withdrawn from
the program. Matrix claims to have removed the terrorism  scoring
system from its final product, but was unable to provide any 
documentation of that when queried by AP. According to documents
obtained  from Utah, which has since pulled out of the program,
Seisent officials  said the scoring system was developed by reverse
engineering an unnamed  "Terrorist Handbook" that describes how
terrorists "penetrate and live in  our society." The scoring
algorithm factored in such data as age, gender,  ethnicity, credit
history, "investigational data," information about pilot  and driver
licenses, and connections to "dirty" addresses used by other 
suspects. (AP 20 May 2004)

(from NewsScan Daily, 27 May 2004)

      A survey of U.S. government agencies has uncovered widespread
data  mining activities, through more than 120 programs that collect
and analyze  large amounts of personal data such as names, e-mail
addresses, Social  Security numbers and driver's license
information. The survey, conducted by  the General Accounting
Office, identified 52 federal agencies that  routinely comb through
citizens' data, and because the GAO figures exclude  most classified
projects, the extent of intrusion into personal privacy  could be
much higher. Sen. Daniel Akaka (D-Hawaii), who requested the 
report, said: "I am disturbed by the high number of data mining
activities  in the federal government involving personal
information. The government  collects and uses Americans' personal
information and shares it with other  agencies to an astonishing
degree, raising serious privacy concerns." An  advisory committee to
the Pentagon chaired by former FCC chairman Newton  Minow, has
recommended that federal agencies generally should be required  to
obtain court approval "before engaging in data mining with
personally  identifiable information" on U.S. citizens. It also
recommended that  agencies should, if possible, strip out all
personal identifying  information before working with such data.
(New York Times 27 May 2004)

More information about the EAS-INFO mailing list