[Corpora-List] Syntactic Parser for German

[Sandra Kuebler]

Hi,

That mostly depends on how you define 'robust' and what kind of  
spoken data you work with. My suggestion would be to use either lopar  
or bitpar. These are trainable parsers, developed by Helmut Schmid  
(Stuttgart) and can be downloaded from the following webpages:

http://www.ims.uni-stuttgart.de/tcl/SOFTWARE/LoPar.html
http://www.ims.uni-stuttgart.de/tcl/SOFTWARE/BitPar.html

Then you could train them on the Tuebingen Treebank for Spoken German  
(developed in Verbmobil, i.e. you get dialog data), which you get for  
free after signing the license from the following webpage:

http://www.sfs.uni-tuebingen.de/en_tuebads.shtml

You would have to write your own scripts to extract the grammar from  
this treebank, though.

Sandra


On Oct 22, 2008, at 9:07 AM, Olga Pustylnikov wrote:

> Hi,
>
> does anybody know a robust syntax parser for German preferably  
> applicable to spoken data?
>
> -- 
> Olga Pustylnikov
>
> Universität Bielefeld
> Fakultät für Linguistik und Literaturwissenschaft
> Universitätsstraße 25
> D-33615 Bielefeld
>
> http://ariadne.coli.uni-bielefeld.de/pustylnikov/
> olga.pustylnikov at uni-bielefeld.de
> _______________________________________________
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora

Sandra Kuebler
Indiana University
Department of Linguistics
Memorial Hall 322
1021 E. Third Street
Bloomington IN 47405
USA
phone: (812) 855-3268
fax: (812) 855-5363
email: skuebler at indiana.edu



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20081022/7de81644/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora