On Tue, 9 Sep 2003 12:50:51 +0200
"Julian Mehnle" <lists@mehnle.net> wrote:
> No, you can't make such a general statement that using content-based filters
> is "better" than using DNS RBLs. It wholly depends on the listing policy of
> the RBL, and in most cases, content-based filters will be the far worse
> option, because it only drives spammers to make their spam stick out from
> the general mail noise less and less! I.e. after prolonged, widespread use
> of content-based filters, spam won't be easily distinguishable from your
> normal mail traffic anymore from a machine's point of view.
I beg to differ. I do not see this happening any time soon. I've seen
some ingenious ways for spam to get through SA's blocks but the Bayesian
classifier has caught them. Like it or not unless the spam is somehow related
to the normal topics of conversation that the individual regularly engages in
and is from places he or she normally gets mail from the classifier is going
to catch it.
The most recent example are the spams now that have maybe 3-4 random words
in them and have the actual ad in an attached or linked image. SA passes it
since it hits maybe 2-3 of SA's markers. However the Bayesian classifier tags
it at 95% or higher. IE, it is being caught by the BC, SA's default
configuration doesn't give enough weight to the BC for that alone to cause it
catch it.
Finally there's the issue that SA is using the RBLs in the manner Karsten
stated, as a weight to determine whether or not the mail is spam and not an
absolute marker that it is or is not spam.
--
Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
PGP Key: 8B6E99C5 | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------
Attachment:
pgp7WC9GcWMX8.pgp
Description: PGP signature