Thanks for the response. My apologies if my message was confusing. We're not using Exchange....the backend is open source running on Linux. We're using Outlook for clients but there is no Exchange integration. I won't be able to copy the messages so I'll have to redirect/bounce them according to the Wiki write-up.
My intention is to create an easy-to-use mechanism in Outlook (using VBA) to allow people to select multiple messages within outlook and choose one of two buttons on their toolbar; one for Ham and one for Spam. Behind the scenes it will redirect/bounce the messages to the appropriate mailbox and at that point I can run sa-learn. I've already got a prototype working in Outlook that just needs some tweaking. The messages will be sent to a mailbox on a Postfix implementation. My question is this: When the message is resent some of the headers in the resent message are different than the original message. Will sa-learn be learning something differently if it analyzes the resent message based on the fact that the headers are slightly different? Thanks. -Jim -----Original Message----- From: Steven Dickenson [mailto:[EMAIL PROTECTED] Sent: Monday, September 13, 2004 2:51 PM To: Jim Ficarra Subject: Re: Bayesian Filtering/Resending from Outlook Jim Ficarra wrote: > I would like to setup a site wide spam filter using SpamAssassin. In > addition to using the network rules, I would like to setup something > where my users can submit their messages for ham/spam to the system so > the Bayesian system can learn. > > I read in the Wiki that you can redirect/bounce a message with mail > headers intact to a couple of mail boxes (one for ham, one for spam) and > run the sa-learn. The following URL in the Wiki > (http://wiki.apache.org/spamassassin/ResendingMailWithHeaders) describes > how to do it for several mail clients. My site uses Outlook. Assuming you're using Exchange, setup two public folders and have users move/copy mail to those folders. Then, use a script on your spamassassin box to read those folders in with IMAP and learn them with Bayes. There've been a few scripts floating around that do this for you. Search the archives for links (not sure if it's in the wiki yet). Steven
