Does anyone have a variation on the IMAP directory trick for learning
spam/ham, modified so that you can drop a message SA identified as spam into
one of the folders, and have the attachment stripped out and cleaned up for
learning?
I just noticed a case that scored lots of points on various rule hits, but
Bayes only gave it 44%. Clearly Bayes should be told a few more of these
are spam, but it is a bit of a pain to clean the message back up from the
Outlook side. A script or some such on the SA machine seems more to the
point.
Loren
