A major study lambasts the US government’s reliance on data mining to detect terrorists as “a waste of resources”.
Since 9/11 the US government has been analysing people’s travel, spending and communications habits, hoping to spot patterns of abnormal behaviour that could lead it to terrorists.
But in its paper, “Effective Counter-Terrorism and the Limited Role of Predictive Data Mining”, public policy research foundation the Cato Institute argues that data mining creates a false positive rate of 90%. Worse still, while the data capture methods are easily avoided by real terrorists, it violates citizens’ privacy and civil liberties.
“The absence of terrorism patterns means that it would be impossible to develop useful algorithms,” points out the report.
“The corresponding statistical likelihood of false positives is so high that predictive data mining will inevitably waste resources and threaten civil liberties.”
Rather than data mining, the report’s authors conclude that more traditional information sharing and investigatory legwork are the answer to curbing the terrorist threat.
I do not question the conclusion of the study mentioned that data mining, *as used* has proven ineffective against terrorism, but I do caution against the broader conclusion which some (Schneier, for one) have drawn that data mining *in general* is ineffective against terrorism. If only for its utility in rooting out money laundering (frequently used to finance terrorist activities), data mining has proven worthwhile.
ReplyDelete