Collect about 200 of 'em or a little more as well as 200 or more ham
messages. Then train Bayes. When Bayes kicks in the results are astounding.
{^_^}

----- Original Message ----- 
From: "Ron McKeating" <[EMAIL PROTECTED]>


> Installed spamassassin on a system yesterday and sent a few test emails.
> It seems to be scoring very low, with lots of rules giving 0.0 score eg
> 
> 0.6 points, 5.0 required) pts rule name              description ----
> ----------------------
> -------------------------------------------------- 0.0
> PORN_16                BODY: Possible porn - nasty, dirty, little etc.
> 0.0 HOT_NASTY              BODY: Possible porn - Hot, Nasty, Wild, Young
> 0.6 SUBJ_ALL_CAPS          Subject is all capitals 0.0
> CLICK_BELOW            Asks you to click below
> 
> Any pointers as to where I should be looking ?
> 
> Ron

Reply via email to