Dear Apertiumers:
Shashwat Goel is one of our GSoC students this year. He is working
(under the mentorship of Jorge Gracia, CCed, and of myself) on software
that generates bilingual dictionary (bidix) entries for a language pair
by inferring them from the relations found in other bilingual dictionaries.
His program has a number of hyperparameters that need to be optimized so
that the output is maximally usable. Currently he generates
lemma/PoS–lemma/PoS entries (no grammatical info yet, just part of speech).
He has generated random samples from predicted output for a number of
language pairs. The samples contain 150 bilingual dictionary entries to
be classified as correct or incorrect (The question is: "would this
translation be correct (in some context)?"). We would like to have 3
evaluators per language pair so that we can deal with ties.
The language pairs are en-ca, en-es, eo-en, fr-ca, oc-ca and oc-fr.
We would be really greatful for Apertiumers to contact Shashwat
(shashwat.g...@research.iiit.ac.in), copying me (m...@ua.es), offering
half an hour of their time to evaluate 150 entries.
Thank you very much!
Best,
Mikel
--
Mikel L. Forcada http://www.dlsi.ua.es/~mlf/
Departament de Llenguatges i Sistemes Informàtics
Universitat d'Alacant
E-03690 Sant Vicent del Raspeig
Spain
Office: +34 96 590 9776
_______________________________________________
Apertium-stuff mailing list
Apertium-stuff@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
