From: "John Andersen" <[EMAIL PROTECTED]>
On Thursday 12 October 2006 14:54, John Rudd wrote:
That rule has a 3.2 value because the 3.2 value is
accurate to differentiating spam vs ham in the corpus. Therefore, the
score is appropriate.
No, its not accurate.
The rule is in-discriminant as to content. It flags ham with the same score
as spam. Therefore by definition it is in-discriminant, and thus useless
as in the prediction of ham vs spam.
Zero that rule's score, and your false positives will fall, but your false
negatives will not increase. The rule unfairly targets ham.
I notice a great influx of spam on monday. Should all mail sent on monday
regardless of content be scored higher?
Post hoc, ergo propter hoc? Is that what passes for analysis?
??jdow?? Run that by me again how it is ineffective? I have log files
that show the RFC rules are reasonably effective except for one of
them, which is not the postmaster rule.
The rule simply declares that a message that triggers the rule looks
like a message that triggers the rule. The scoring system acccepts a
score attached to that fact about the message. The score is generated
by measuring how good the rule is at catching spam as opposed to ham.
This is true of EVERY rule. Some of the scores may be wrong. But the
rules are not.
Your mind seems to have taken a vacation or something, John. Calm down
and think it through. You are saying that EVERY rule is unfair when you
say RFC postmaster is unfair. They ALL operate on EXACTLY the same
principle. Generate a hypothesis about messages that you hope is more
or less true about spam than about ham. Then a score is given that
rule after testing it against data that hopefully resembles live data.
Somebody tested the hypothesis that email from broken postmaster sites
might be such a distinguisher and tested it. It was. A value was given
to the rule. How is this in any way difference from a rule that looks
for the colloquial word for fornication or its obfuscated equivalents
as a possible distinguishing feature?
Calm down and reanalyze. Then vent on the abject idiots who refuse
receipt of email based on only ONE criterion test from a site that
has a bad attitude, such as Paul Vixie's old black hole list. (If I
EVER am face to face with that man he's earned a kick in the "sweet
spot". The good he has done for the community mitigates this to a
really hard slap on the face. He cost me some money with his broken
blacklist.)
Also feel free to attack sites that simply run a few anti-spam tests
then toss or refuse email that fails the tests rather than performs
a SpamAssassin type markup so the user can filter and vet it for
errors. Then if the site fails to monitor the ACTUAL error rates
for their anti-spam and work to minimize the false alarms give the
administrator the "kick in the sweet spot" he has earned.
{^_^}
