> I tried sa-learn, but don't you need a sizable spam collection for it to > work? The docs say that you need to collect about a thousand of ham and > spam messages before the training starts to work. That sounds like a > pain in the neck. Or am I missing something?
By default you need 200 hams and 200 spams before Bayes will kick in. You should continue to feed it for a while with choice morsels, and feed it anything it gets wrong so that it can correct its ways. (Also feed it spam it scores around Bayes_50, which indicates that it doesn't know that it is spam.) Loren