don groves wrote:
> At 14:22 8/11/2004, don groves wrote:
> >At 13:29 8/11/2004, Brandon J. Van Every wrote:
> >
> >Maybe if you spelled S*attle correctly the filter would pass it.
> >Most, if not all, words with embedded non-alpha characters these
> >days are ads for V!agr@ or C!@lis so the filter gets trained
> on those.
>
> Nope, that's not it ;)

I am unsure of your antecedant, so I'm unsure what you mean by your
followup.

If you look at the X-Sp*m header of your 2 posts, you will see a list of
words.  Those with high values like 99's increase the likelihood that
your post will be classified as sp*m.  That is why I use an asterisk
with them.  The problem words in your own posts are:

  brand*n
  spell*d
  s*attle
  agggggr (only 1 g, probably part of vairga (letters transposed to
protect us all))
  addddds (only 1 d)
  traff*c

Bear in mind that each and every one of my posts carries an automatic
strike against me.  Even though I'm willing to use Brand*n in my
signature, I'm totally unwilling to create a separate e-mail identity
just for caml-list.  So every one of my posts has 1 correctly spelled
occurrance of Brand*n in it.

I bet, furthermore, that you will not see the word S*attle (correctly
spelled) in the X-sp*m header of this post.  This would disprove your
claim about * being a red flag for the filter, unless the filter does
not in fact report full information about how it makes its decisions.


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners