On 09/23/12 18:28, John Hardin wrote: > On Sun, 23 Sep 2012, James wrote: > >> I wrote this little script to update the bayes rules. I can do this on my >> imap account but my pop3 account gets way more spam and the messages are no >> longer on the machine with sa once I pop them off. >> >> Any comments on my script? > > Bear in mind that spams which don't score high enough to be quarantined or > discarded will end up in your inbox, as will false negatives. Training all of > the mail in all of your inboxes as ham will train these messages as ham and > make any small error in classification much worse over time. I will manually move spam to an Junk (not a .INBOX* > > During the initial training period you want to manually review messages and > build a ham corpus and a spam corpus. Once bayes is running you generally > only want to train on misclassified messages. This decisionmaking process > cannot be automated, or the errors wouldn't occur in the first place. > > You should set up per-user train-as-ham and train-as-spam mailboxes, and only > train from those, only for the users whose judgement you trust. Then, those > users should copy misclassified messages to the appropriate folder and may > also add samples of ham to the train-as-ham folder whenever desired.
The only user is me. :-) Is there a way to convert my Thurderbird bayes to spamassassin? > >> >> #!/bin/bash >> >> IFS=$'\n' >> FOLDERLIST=`find Maildir -name .INBOX\* -type d;` >> >> for i in $FOLDERLIST; do >> echo "Processing ""$i" >> # `sudo sa-learn"--ham "$i"` >> done >> >> #`sudo sa-learn --spam Maildir/.Junk >> > >
