Hello there (I'm pretty newbiew on SA) and maybe my questions are stupid, but any way,
1. When I run SA-Lear on HAM and SPAM confirmed directories (let's talk about 800 mails on each one), but after that I still getting some SPAM that aren't detected, If I wan to run SA-learn again, do I have to run on a complete directory (old messages and new ones) ? or can I run sa-larn on a new mails only ? (SPAM and HAM obviously) 2. Once I ran sa-learn, do I have to make changes on .cf ?, What I did, is just install SA, install EventSynk for Exchange and then try to train it and thats it what I did, do I missing someting ? 3. (Last one) I put this on my local.cf whitelist_from [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] whitelist_from [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] whitelist_from [EMAIL PROTECTED] [EMAIL PROTECTED] whitelist_from [EMAIL PROTECTED] whitelist_from @champlabs.com whitelist_from @bordercustoms.com Is that valid ?, I mean, can I put as many addresses as I need on whitelist_from ?, this is becouse SA are tagging mail as SPAM from those people, and they are part of the company. Hope I explain my self (my english isn't good) Thanks in advance. =========================================== Efren Pedroza Huerta -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Robert Menschel Sent: Thursday, December 11, 2003 8:47 PM To: Graham Borland Cc: [EMAIL PROTECTED] Subject: Re: [SAtalk] Training based on previously detected spam -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello Graham, Thursday, December 11, 2003, 7:21:52 AM, you wrote: GB> I'm about to start training sa with spam and ham. Is it a good or bad idea GB> to use spam which has already been detected by sa's non-trained tests, or is GB> that a waste of time? Should I only train it with spam which is currently GB> managing to slip through? You should train with all of your confirmed ham, and with all of your confirmed spam, regardless of SA's previous results. If you want, you can separate out those messages which were a) correctly identified, and b) went through the Bayes auto-learn stage, but that isn't necessary. Bob Menschel -----BEGIN PGP SIGNATURE----- Version: PGP 8.0.3 iQA/AwUBP9ksDpebK8E4qh1HEQKNTgCfVLoHbT2MAFmWhpHMrMV/4RCQ7eAAn0ev mQcGYgBaXMTPc4mZ7SchDhQH =ZaEQ -----END PGP SIGNATURE----- ------------------------------------------------------- This SF.net email is sponsored by: IBM Linux Tutorials. Become an expert in LINUX or just sharpen your skills. Sign up for IBM's Free Linux Tutorials. Learn everything from the bash shell to sys admin. Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk ------------------------------------------------------- This SF.net email is sponsored by: IBM Linux Tutorials. Become an expert in LINUX or just sharpen your skills. Sign up for IBM's Free Linux Tutorials. Learn everything from the bash shell to sys admin. Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
