What a surprise. The basic problem is false positives and false negatives, innocent people who are tagged as terrorists and terrorists who are tagged as innocent people. Let's say your city has a population of one million people. If it's 99% accurate, it identifies 1000 people as terrorists, it also identifies 1000 terrorists as innocent people. Do you think it's 99% accurate? I'd say it's probably not.
Still, I doubt this will be the stake through the heart of trying to catch terrorists through data mining.
http://blog.wired.com/27bstroke6/2008/10/data-mining-for.html
Still, I doubt this will be the stake through the heart of trying to catch terrorists through data mining.
http://blog.wired.com/27bstroke6/2008/10/data-mining-for.html