LFG systems and LFG computer implementations

[Lori Levin]

Stefan et al.,

Do any of the implemented LFG systems use statistical methods or machine 
learning?   Also, is there any machine learning or grammar induction 
from treebanks?  Just curious.


--Lori


Lori Levin
Language Technologies Institute
Carnegie Mellon University



Martin Forst wrote:
> Hi Stefan,
> 
>> • Deutsch (Rohrer: 1996, Berman: 1996, Kuhn und Rohrer: 1997, Forst:
>> 2006, Forst und Rohrer: 2009)
> 
> There are probably more (and better) references for the German ParGram 
> LFG. An important one is Dipper (2003); another one, which gives a good 
> overview of the state of affairs at the time, is Rohrer & Forst (2006).
> 
>> • Spanisch
> 
> There is a company in Seville that seems to have their own Spanish (and 
> English?) grammar(s) as well as their own processing platform. They were 
> at a ParGram meeting a few years ago, which I missed, so I don't know 
> any details or references.
> 
> Apart from that, I have a little Spanish grammar, but it's very 
> phenomenon-driven, not broad-coverage by any means and not documented 
> anywhere, let alone in publications.
> 
>> • Vietnamesisch
> 
> As far as I know, the ParGram community hasn't heard from our Vietnamese 
> partners in a long time. Not sure how far that grammar has ever come 
> along...
> 
> I hope this helps.
> 
> Schöne Grüße
> 
> Martin