On Tuesday 31 August 2004 09:32 pm, Rob Blomquist wrote: > > Rob, you have to feed sa-learn two different bunches of mail. > > One known spam and the other known ham. You have to continually > > train it with missed spam/ham. > > > > I do this by having users create two separate maildirs in their Mail > > directory, one named NotSpam and the other named MissedSpam. > > > > Then I run the following script nightly via cron. I had to hack it > > a bit because the original was for mbox, not maildir > > > > #!/usr/bin/perl > > Nice script. I would love to try it out, but what I do is run it against my > mailboxes, and sa-learn skips messages it has sent. > > How could I modify it to replace my poorly written script? > > echo "SpamPile" > sa-learn --spam ~/.Mail/SpamPile/cur/ > echo "inbox" > sa-learn --ham ~/.Mail/inbox/cur/ > echo "Ads" > sa-learn --ham ~/.Mail/Ads/cur/ > echo "VFB' > sa-learn --ham ~/.Mail/VFB/cur/
Well you will note that the script says nothing about how spam and ham gets into the spam and misses spam maildirs. This depends on the MUA the users use. I use Kmail and can simply drag spam to the missed spam box it it will do it correctly. Other user agents hork up the mail when doing this such that it might not be recognized. I don't find I have to feed it all the mail on a continuing basis. (Thats what auto learn is supposed to do). Just the mail it misses, FPos and FNegs. Why on earth would you want to learn Ads as ham? a -- _____________________________________ John Andersen
pgp0wqZfiAjbz.pgp
Description: signature
