"Universal spell-check" algorithm and function morphemes
"Universal spell-check" algorithm and function morphemes
- Subject: "Universal spell-check" algorithm and function morphemes
- From: Jim Witte <email@hidden>
- Date: Wed, 29 Jan 2003 22:56:59 -0500
I notice that in a paper if I write 'hyperglobalist' and
'hyperglobalization' I have to tell the system to learn both of them.
The same thing happens with words like 'blog' and 'blogging'. Does
anyone know (can they hint, or can they suggest to those who would be
able to do something about it) if there are plans to extend the
spell-check algorithm to deal with such functional morphological
differences? In English, I don't think this would be too hard (at
least not a system that would recognize 80% of all morphological
variants that it should) because English function morphemes are often
at the end of words. Although in other languages it could be *quite* a
bit more difficult, especially languages like Hungarian that have
infixes for almost everything (subject pronoun, object pronoun,
preposition, diminutive, etc)..
I seem to remember something somewhere about an linguistic analysis
toolkit or NLP system of some kind that was part of the SDK package at
one point. Was this back in the days when Plaintalk actually used
grammars and such (maybe it still does..), or am I (wishfully)
imagining things?
Jim Witte
email@hidden
Indiana University Computer Science
_______________________________________________
cocoa-dev mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/cocoa-dev
Do not post admin requests to the list. They will be ignored.