From: Eric Hodel <drbrain@...7.net>
Date: 2003-02-21T01:55:33+09:00
Subject: Re: optimization question


--hK8Uo4Yp55NZU70L
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Travis Whitton (whitton@atlantic.net) wrote:

> > Since I've lost the original email, I'll just say that I there's a
> > more efficient way to do what you're doing.  Keep a real database of
> > tokens, and use mail to update it, rather than re-tokenizing the whole
> > email set.
>=20
> This is a good idea, and it looks like an interesting project. Does it al=
low
> you to create a database file from already existing corpus files, or does=
 it
> require you to start from scratch? I've spent a good amount of time build=
ing
> up the 2000 some-odd messages I'm currently using to gather my statistics,
> and I'd hate to have two huge corpus files wasted. Also, the main reason I
> started this project was to learn, so I'd still like to find a fast way to
> update a hash key without having to access it twice.

Yeah, I understand the learning bit :)  I do lots of stuff that other
people have done, just for the hell of it.

You'll just feed each corpus file into the filter, one marked as spam,
the other marked as good.  After that, you're done.  Use procmail to
look at each message, decide if it is spam or not, then place it in a
spam folder if its spam, or let it go through the rest of your filters
if not.  There's also a mutt keybinding that marks it as spam/good mail,
which removes it as spam, then re-adds it as good.

Also, if you want help with RubyInline, I suggest you mail
ruby@zenspider.com, or the author himself, he doesn't read this list.

--=20
Eric Hodel - drbrain@segment7.net - http://segment7.net
All messages signed with fingerprint:
FEC2 57F1 D465 EB15 5D6E  7C11 332A 551C 796C 9F04


--hK8Uo4Yp55NZU70L
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQE+VQiCMypVHHlsnwQRAkcuAKDFbQfrhmRyWIitPDWW/7vfNaMEYQCfeqwK
s0cIts7/tOcCdxu83ufxHSg=
=ExmA
-----END PGP SIGNATURE-----

--hK8Uo4Yp55NZU70L--