From: "Craig McLean" <[EMAIL PROTECTED]>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Robert Bartlett wrote:
Since finding out the trusted_network issue I question the rest of my
local.cf setup. Right now I have AWL turned off and auto learning for bayes
turned off. My question is does SA benefit from turning those 2 back on? Of
course I would clear out AWL and bayes and start from scratch if I did. But
would it make it easier for bayes to be "poisoned" if I turned auto learn
on? Im on SA 3.0.1.
...
Bayes: Round here, BAYES_99 hit 92.77% of spam and 0.40% of ham in the
last year, BAYES_00 hit 80.31% of ham and 0.41% of spam in the same
period. I'd say a well trained bayed db is *very* worthwhile.
As for "bayes poison", I don't think it exists in the sense it's being
used here. Almost every spam I've got with long, random paragraphs in it
There is no inherent Bayes poison that can be placed in spam to poison
the database. However, sufficiently large numbers of mistrained messages
amount to leaving the Bayes database in a poisoned state that leaves it
basically ineffective. In that case it's better to toss the bad Bayes
training out and retrain from the start.
{^_^}
