I think that Timothy is after a descriptive framework for Greek (a set
of categories and their labels), expressed in XML; these can of course
be implemented within a TEI document (most probably as attributes of
different elements). I am aware of the work conducted in Perseus for
a Latin Treebank:
http://daphne.perseus.tufts.edu/docs/guidelines.pdf
Have also a look here:
http://gircse.marginalia.it/~passarotti/
Perseus has also developed an automatic, rule-based morphological
analysis, called Morpheus, which you probably are aware of. Some
detail can be found here
http://www.perseus.tufts.edu/Articles/exploration.html
The TLG is currently also developing its own morphological parser/
lemmatiser, the main engineer as far as I know is Nick Nicholas; you
might want to contact him for more details.
I hope this helps.
Best wishes,
Notis
On 28 Jan 2008, at 09:55, Garces, Juan wrote:
> Dear Timothy,
>
> I fully concur with James. Using an established standard, such as
> TEI (I have to confess my bias!), has many advantages in terms of
> general support (e.g. the TEI discussion list), possible data
> exchange with other similar projects, longevity of data, and
> application of tools.
>
> Some projects, such as OpenText.org, probably feel strong enough to
> develop their own schema instead of using or adapting existing ones.
> There are ways of dealing with such a situation in relation to the
> above mentioned issues, but I would rather prevent such a situation
> in the first place.
>
> Another advantage of using an established schema such as TEI is that
> many texts are being/will be made available in TEI already (e.g. by
> Perseus). Adding a further layer of linguistic markup would be much
> easier, if done in TEI too.
>
> If you are interested in adding linguistic markup with TEI, Notis
> Toufexis (http://www.mml.cam.ac.uk/greek/staff/nt262/) - I hope you
> don't mind me mentioning you, Notis! - has done some interesting
> work with Mediaeval Greek and I seem to remember him recommending
> TEI...
>
> All best,
>
> Juan
>
>
|