[Corpora-List] Machine Translation and Spelling Correction

Francis M. Tyers ftyers at prompsit.com
Fri Dec 4 14:55:10 UTC 2009


El dv 04 de 12 de 2009 a les 09:48 -0500, en/na Alexander Murzaku va
escriure:
> You should definitely look at this
> page: http://norvig.com/spell-correct.html
> 
> 
> I have worked for many years with spell-checkers that have been or are
> OEM-ed by all the biggest software providers. However, I have found
> the information in this site quite complete and up-to-date. Beside the
> 21 lines of Python code, it has links to most applications and
> theories on spell-checkers.
> 
> 
> I would only add the use of FST/FSM based dictionaries and
> morphological engines (Xerox PARC/Inxight and Teragram) in
> spell-checkers.
> 
There are also free software projects (HFST[1] and Foma[2]) that are
drop-in replacements for the proprietary Xerox FST compiler toolkit.

Also, for morphologically complex languages, the Voikko library/tools
might be useful: http://voikko.sourceforge.net/

Fran


1. http://www.ling.helsinki.fi/kieliteknologia/tutkimus/hfst/
2. http://foma.sourceforge.net/


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



More information about the Corpora mailing list