Started Training: to get the 65/35 Spam/Good ratio Filtered mail shows 4015 Good messages & 23 Spam. This ratio should be reversed. I assume that the spam is from Mail’s or my classification. However the almost all of the spam is in Suspect Mail. Is there a way to tell this to Spam Sieve?
Thanks
The recommended ratio is for the Corpus section of the Statistics window, not the Filtered Mail section. The Corpus numbers are based on messages that you’ve manually trained as well as messages that SpamSieve auto-trained. Suspect Mail sounds like a special mailbox that’s used by your server junk filter. You can certainly train some messages in there as spam if needed. But note that a few hundred messages total is plenty for the initial training.
Thanks for clarifying this, Michael