[Debtags-devel] Using Python NLTK for tag generation [was: AI for tag generation].

[Enrico Zini]

--rwEMma7ioTxnRzrJ
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, Oct 03, 2004 at 06:06:46PM +0200, Herv=C3=A9 Eychenne wrote:

> > Tag: special::completely-tagged
[...]
> >  This tag will also be automatically removed from all packages after
> >  vocabulary updates.
> Tagged carefully, ok... but at which time?
> What if some new tags appear after the review, making the tag set of a
> package not so up-to-date anymore?

Well:

> >  This tag will also be automatically removed from all packages after
> >  vocabulary updates.

Of course if I perform a vocabulary restructuring, I may be nice and go
through all completely-tagged packages bringing them up to date, so as
not to fully invalidate previous efforts.

> In my opinion (I said that from the begining, if you look at my first
> posts), we should take care of that. I was proposing that new tag
> additions (addition of a new tag name, or addition of a tag in a package
> taglist) should be timestamped.

I'd say timestamped, and keeping track of who made the update.

Or, tagging each package with a tag which contains the version of the
vocabulary at the time of the tagging.

But this all would need serious restructuring on the server
infrastructure.  So far, this is probably the move with the best
outcome/effort value, and it will make it possible to start
experimenting with bayesian things.

In a future refactoring, if we will have seen that this all is actually
used, we bring everything in.


Ciao,

Enrico

--
GPG key: 1024D/797EBFAB 2000-12-05 Enrico Zini <enrico@debian.org>

--rwEMma7ioTxnRzrJ
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFBYD4r9LSwzHl+v6sRArqmAJ9yEP0wnQagWhZY7pNRg/C2KHQspgCgiOC8
5HPY4SRqg/KzmMXOUN7WGP0=
=1X5b
-----END PGP SIGNATURE-----

--rwEMma7ioTxnRzrJ--