[linux-elitists] X-AmikaGuardian-Category

Nathaniel Smith njs@pobox.com
Sat Oct 18 16:12:03 PDT 2003


On Sat, Oct 18, 2003 at 05:58:36PM -0400, Matthew W. Miller wrote:
> Am I the only one who finds some of thsese 'AmikaGuardian' headers
> amusing?  How does it spot a 'Pastime' or a 'Joke'?  The Amika people
> must have one mother of a Bayes system scanning every single e-mail
> message on the planet.  I'd be sorely disappointed if it wasn't.

I guess you're sorely disappointed.  Your basic naive Bayes
classifiers are a pretty weak technique; the state of the art in text
classification and natural language data extraction is way, way beyond
that.  There's quite a literature on this sort of thing.  (Partly
because there's lots of government funding for it.  The main problem
facing organizations like the CIA is not gathering information, but
rather sorting through the incredible amounts of information they take
in on a daily basis... but hey, at least M-x spook is still fun, if
not terribly effective.)

-- Nathaniel

-- 
So let us espouse a less contested notion of truth and falsehood, even
if it is philosophically debatable (if we listen to philosophers, we
must debate everything, and there would be no end to the discussion).
  -- Serendipities, Umberto Eco



More information about the linux-elitists mailing list