[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Spamassassin + exim



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sat, Aug 30, 2003 at 08:27:43PM -0700, Steve Lamb wrote:
> > Did I miss something?
> 
>     How does it get to that 200 mark of ham and spam?  

I got bored last night.  Really bored.  Waiting for 2.6.0-test2 to
compile bored.  In the mean time, I went through my email archive.  It
goes back to last March.  I sorted it into spam and ham.  I then used
sa-learn --showdots --ham --mbox ham & to learn all the ham, and
sa-learn --showdots --spam --mbox spam & to learn all the spam and
watched a few thousand dots scroll down the screen.  I wasn't thinking
ahead or I would have used time on it to see just how long it took (I
went and did other stuff for a while.  I was bored.  I watched a lot
of Star Trek off the Tivo and had a Proper Cup of Tea).

It learned 542 spam.  It learned 14,000 ham.  It took forever.  I wish I
used time but I'm not starting it again.  It didn't help that I was
compiling 2.6.0-test2 in the background.  It's getting warm in the
apartment, so the fans on my koolance case kick up a notch every so often[1].

>     sa-learn lets you feed messages to the Bayesian filter to learn from. 
> Bayesian doesn't need to be active for it to learn.  This lets you get the
> filter trained a bit faster.  Furthermore it will help adjust on messages that
> SA would miss as either ham or spam.  The bounces and virus messages being a
> prime example.  Ever since they were coming in I've been feeding them to the
> filter.  Now most are rejected at SMTP time even though the SA team has not
> released another version with updated filters to address those types of
> messages.

Woohoo!  <zim> Another win for the Debian army! </zim>

>     I don't sent every message to the filters.  I let autolearn do its job. 
> But I do make it a point every now and again to feed it 20-30 messages from
> random lists and my inbox.  About 1/2 were not autolearned so it keeps my
> filters fresh.  I do feed all spam into the filters to ensure that side is
> definitely kept up to date.

I added this little bit to my ~/.muttrc:

macro index   [	       "|sa-learn --ham --single 2> /dev/null > /dev/null &<enter>"
macro pager   [	       "|sa-learn --ham --single 2> /dev/null > /dev/null &<enter>"
macro index   ]	       "|sa-learn --spam --single 2> /dev/null > /dev/null &<enter>"
macro pager   ]	       "|sa-learn --spam --single 2> /dev/null > /dev/null &<enter>"

Wham, every message can be easily run through.  Autolearn becomes
handy when you're on vacation (since it'll be extremely tuned).


[1] The case is reminiscent of a locomotive if you look at it at the
right angle.  I'm thinking a cool case mod would be to add vents and
and "cab windows" to make it look like one.  Hey, it's already got the
big radiator and fans at the top.

- -- 
 .''`.     Paul Johnson <baloo@ursine.ca>
: :'  :    
`. `'`     proud Debian admin and user
  `-  Debian - when you have better things to do than fix a system
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQE/UlblUzgNqloQMwcRAprNAKCwE/0eCObF/PeH73qfbYmryvjztACgi87k
aI599imaCLqbLgJt1VITXnI=
=3C/a
-----END PGP SIGNATURE-----



Reply to: