Bayes’ theorem(probability) to combat spam mails..

The other day I was reading some stuff on conditional probability and Bayes’ theorem and my mind wondered what could be the possible application of this theorem apart from the usual ones pertaining to manufacturing and alike.

Digging a little, I came across this interesting piece on the web which refers to Bayes’ theorem to limit or restrict spam mails in the inbox. But before I go further, a little bit on Bayes Theorem.

It is “A theorem stating that the probability of a hypothesis, given the original data and some new data, is proportional to the probability of the hypothesis, given the original data only, and the probability of the new data, given the original data and the hypothesis. Also known as the inverse probability principle. “
( http://www.answers.com/topic/bayes-theorem )

So, basically, the logic would work on the above lines by trying all the possible combination’s from a percentage of the mails ( which would be a subset) having a keyword of the spam and apply it for those which have the word or are related to it.

Interesting..? then watch the video for more..

More information on Bayes’ theorem can be found at wikipedia