On Tue, 2009-01-06 at 19:31 +0100, Kai Schaetzl wrote:
> It appears to me that the HABEAS rules are hitting only a very tiny fraction
> of
> mail, many of the nightly mass-checks don't have a hit at all (or is it that
> those
> checks don't contain any network checks?). The aggregated view shows no hits
> at all
> for these rules.
Network tests are done once a week, not daily.
> I'm not sure if I'm reading the ruleqa correctly, although I read it's help.
> 1. I'm wondering why many rules show a score of 0.0
These appear to be network tests.
> 2. do I understand it correctly that a nightly check contains only the spam
> received over the last 24 hours?
No. The nightly mass-checks contain the full corpora.
> 3. I don't see any explanation for s/o and rank. (Rank seems to be some sort
> of
> ranking according to the hit rate, but I find it hardly understandable that a
> rule
> that hits a lot of messages, like URIBL_SURBL, scores 1.0 as rank and a rule
> that
> hits almost no messages still scores at half of that. s/o seems to show the
> ham/spam ratio cleanliness?)
Correct, S/O is the Spam / Overall ratio. The higher that ratio, the
better the rule and the lower the ham hits (in percent, not absolute
numbers).
--
char *t="\10pse\0r\0dtu...@ghno\x4e\xc8\x79\xf4\xab\x51\x8a\x10\xf4\xf4\xc4";
main(){ char h,m=h=*t++,*x=t+2*h,c,i,l=*x,s=0; for (i=0;i<l;i++){ i%8? c<<=1:
(c=*++x); c&128 && (s+=h); if (!(h>>=1)||!t[s+h]){ putchar(t[s]);h=m;s=0; }}}
