Training the Spam Terrier. Simply specify folders containing valid and spam messages and let the Bayesian filter analyze their contents to determine the unique characteristics you use to define spam. The more you train the filter, the more accurate the spam identification results will be.
