Matt Kettler <[EMAIL PROTECTED]> writes on 29 August 2003 at 18:37:38 -0400 > At 03:13 PM 8/29/2003 -0500, David Dyer-Bennet wrote: > > >I notice that there are different scores for tests depending on "how" > >spamassassin is running, and that one of the differences is whether > >Bayes filtering is turned on. But I've explicitly set use_bayes, and > >yet I never see a Bayes score turn up in the summary on a message. > >I've put a bunch of messages through sa-learn (or spamassassin > >--report), and the Bayes files I configured have grown and their dates > >got updated. What else should I be checking? > > 1) did you punch any non-spam (ie: ham) messages into sa-learn, or just > spam? SA must learn at least 200 spam, and 200 ham messages before it will > use the bayes engine.
I believe you have nailed it. I was adding some more on general principals, and it started producing results occasionally. Also there are a number of middle Bayes ranges that have score zero associated with them by default, hence they don't show up in reports, so I had to try some messages that were more clearcut. -- David Dyer-Bennet, <[EMAIL PROTECTED]>, <www.dd-b.net/dd-b/> RKBA: <noguns-nomoney.com> <www.dd-b.net/carry/> Photos: <dd-b.lighthunters.net> Snapshots: <www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera mailing lists: <dragaera.info/> ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
