html comment processing

[Greg Louis]

On 20030330 (Sun) at 0912:55 -0500, David Relson wrote:
> At 07:58 AM 3/30/03, Greg Louis wrote:
> 
> >and make sure there's an improvement.  I've got two more s/mindev scans
> >going at present, but after that I can find clock cycles for this
> >purpose.

> Of course, an experiment would be of value.  Can you crank up your fast 
> machine?
> 
It should be finished the current run by Tuesday night, I hope.  What I
don't know is whether I've actually got spams in the kitty that contain
the bogus tags, nor how many.  It would be convenient to create a
corpus consisting of nothing else, split it between training and test
in the same proportion as the rest of the spam, and then see how well
the strict processing measures up against the loose.  Has anyone a
supply of these that they could zip up for me to fetch with anon ftp?

I too found 0.10.1 to work well, but the population of spam has changed
somewhat since then (more circumvention attempts), so I don't know how
relevant that finding is.

-- 
| G r e g  L o u i s          | gpg public key: finger     |
|   http://www.bgl.nu/~glouis |   glouis at consultronics.com |
| http://wecanstopspam.org in signatures fights junk email |