Re: shrinking css files for quicker training?

[email protected] Tue, 18 Dec 2012 10:00:27 -0500
Newsgroups gmane.mail.spam.crm114
Message-ID <[email protected]>
Kurt Bigler <kkb-qsUyPSV3HvqUK90frp/[email protected]> writes:

> Hi, all,
>
> I've been mostly just letting my crm114 mailfilter installation just "sit"
> for a number of years and it just keeps on working, of course.
>
> However, along the way I think I may have expanded my css files too much,
> perhaps too concerned about packing density.  I judge this from the fact
> that the learning seems entirely too sluggish compared to what it used to
> be.  

Hmmm... try cssutil -r <cssfile>  and post the results here.

It may be the case that you do indeed have too much stuff in there and
the relay chains are just getting too long.

> I assume so much has been trained by now that new training has less than the
> desired impact.  This may have worsened due to the number of years the css
> files have been in use, and just how much spam has changed over that time.
> I would like some of the old training to "fall off the end" but I don't have
> the time required to just start over right now.

Ahhh- "aging" - the Castle Anthrax of text recognition.

That is a problem, which is *not* circumvented by any system I know of.

All of those wonderful examples of spam from five years ago are
completely obsolete, and possibly misleading.

My advice is (sadly) that you can't really "age" it - the easiest 
thing to do (which will at least give you more space and make too-long
chains shorter again) is to use cssmerge to create a bigger file
(disks are bigger now, dude!)

Other than that, you can start again.

Some (a few) versions of cssutil also had the ability to 
zeroize small bin values.  That might be useful at least for
speedup- however, accuracy will suffer as you are lobotomizing
out low-count samples.

Make a copy of your .css files before you do that!

Failing that, you're in for a retrain.  :-(

   - Bill

------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d