[Corpora-List] Stemmer for Hindi

[Daniel Zeman]

Dne 18.11.2010 13:53, Francis Tyers napsal(a):
> ...
> You could try the morphological analyser from IIIT if you can manage
> with the WX notation:
>
> http://ltrc.iiit.ac.in/showfile.php?filename=onlineServices/morph/index.htm

Thanks for the link, Fran. We too have been trying to locate a Hindi MA 
for translation experiments. We got something from Hyderabad a year ago 
but we did not manage to get it running then. Their current version is 
much better - great thanks to IIIT!

For Luís and others, I have a convertor from UTF-8-encoded Devanagari to 
WX and back. I also rewrote the spell_variation.pl script because the 
original version was incredibly slow. Contact me if you are interested.

Dan


_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora