El 2018-07-07 15:01, Hèctor Alòs i Font escribió:
The new pair fra-oci is ready for a 0.1 release. It is been developed
by Claudi Balaguer as a GSoC 2018 project. The translator from French
to (Reference/Standard) Occitan should be now published in order to
easily get feedback.
. The bidix has almost 37500 entries
. There are 120 t1x rules, 12 t2x rules and 1 t3x rule
. There are 851 lexical selection rules
. The French morphological disambiguation has been improved: it has
now 611 rules
. The "naive" coverage in a Wikipedia corpus is 92%
. WER is around 13% in Wikipedia articles
. There are not any testvoc errors (from French to Occitan)
Of course all this is mostly the result of working with two
close-related languages on which there has previously been a lot of
work in Apertium, in addition to extensive work on the tools that make
up the Apertium platform. Thanks to all!
In the next weeks we will improve this version, while we work on the
opposite direction (from Standard Occitan to French).
@Tino, could you prepare a release, please? The related language
repositories are: apertium-oci-fra, apertium-oci (not used in a
released pair until now), and apertium-fra.
Besides, apertium-oci-fra should be labeled as "trunk".
Excellent work, thanks Hèctor and Claudi!
Can we also work on getting this into the Wikimedia Content Translation
too ?
As soon as we get it enabled on the main Apertium site, we should do a
press
release or publish it to various Occitan groups. :)
Thanks,
F.
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Apertium-stuff mailing list
Apertium-stuff@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
