On Thu, Sep 22, 2005 at 10:59:57AM -0400, Louis Proyect wrote:
I have a literary account of my struggle with base64 spam at:
http://unrepentant.blogspot.com/
It prompted this comment:
You might want to look at moving away from Procmail recipes
and SpamAssassin to a pure Bayesian filter [PG's Plan For
Spam]. Base64 encoding becomes a spam-indicator if you get Base64
spam and no legitimate email Base64 encoded.
If you want a working version (with Bayesian noise reduction!)
check out DSpam.
(Full disclosure: I did a degree dissertation and project on
Bayesian filtering for spam killing).
Max | Email | Homepage | 09.22.05 - 9:00 am | #
While I have nothing at all against Bayesian, and think it's
cool in general, I believe it does need a fair amount of room
for the sampling files. But anyway, all I really want to say
here is, it is not difficult to find base64 stuff with procmail.
I do it in my own .procmailrc, and it's not at all one of my
more complex or nuanced recipe sets. What I have coded in
has a whole bunch of private dependencies (private vars), however,
so I can't all that easily just copy and paste it in here for
your consumption. If I have time at some point and nobody else
does it first, maybe I'll get around to posting a sample solution.
As for using the trait as an indicator, yes, of course. That's
a prime use of scoring in procmail. And Sean Straw talks about
it often and gives examples of how he ammasses and scores these
indicators. I think he has samples on his procmail pages, too,
with a URL found in his .sig.
I don't use Bayesian because (a) I don't need it, having
a very nice working solution in procmail-only; (b) I am
very tight on space and don't have room for the sampling libraries
on my shell account; and (c) I was pretty much all set with
my procmail stuff before it, and also SpamAssassin before it,
came out.
--
dman
____________________________________________________________
procmail mailing list Procmail homepage: http://www.procmail.org/
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail