[USFRIENDINVICTORIA]

Here's what I posted earlier on another thread:

"Let's assume, for the moment, that this story is factual and complete. If so, it illustrates one of the (many) problems with comprehensive electronic surveillance, combined with data mining -- False Positives.

If the DHS, FBI, etc. had prior reasonable cause to suspect this couple, and that caused them to check their Google searches -- then, those results might be considered evidence.

However, when you surveil everyone routinely, a large number of apparent hits will occur randomly. (Think of Google search topics as cards -- deal enough of them, and you will have a number of random "hits" of "hot" topics.) Anyone, who has studied (even somewhat) advanced statistics, will understand this problem quite well. Rather than using the data mining to confirm suspects -- they're using data mining to produce suspects. Given that there are probably a lot more people, who happen to search on two or more trigger words, than there are actual active terrorists, at any given time -- you are going to generate more false positives than real hits. A lot of innocent citizens are going to be treated like terrorists; because of this bass-ackwards way of collecting evidence."

http://www.freerepublic.com/focus/f-news/3049950/posts

It seems that the story was not complete, nor completely accurate. However, the rest of what I said still applies.

In short: I agree with you.

[dirtboy]

Well said. I work with marketing databases. I have said for some time that if we really want to find terrorists in this country, turn the data over to marketers. They would have them segmented into the exact variety of terrorist in three weeks.