Group: http://groups.google.com/group/link-grammar/topics
Linas Vepstas <linasvepstas@gmail.com> Aug 16 01:12PM -0500
^ > One project that has never been done, and is fairly good for a
> beginner I think, is to try to test how well RelEx works relative to
> other parsers...
[...]
> identical to RelEx's, but I
more...
Ben Goertzel <ben@goertzel.org> Aug 16 11:18AM -0700
^ Interesting, Linas...
So, my suggested project for the student interested in an NLP project,
was to use some scripts to transform a large corpus into dependency
form (tweaked to agree with
more...
Linas Vepstas <linasvepstas@gmail.com> Aug 16 01:42PM -0500
^ > was to use some scripts to transform a large corpus into dependency
> form (tweaked to agree with Stanford output format, say), and then
> test RelEx vs. Stanford on this large dependency
more...
Ben Goertzel <ben@goertzel.org> Aug 16 11:47AM -0700
^ Thanks Linas, that is a good suggestion
So for G Meera: perhaps Linas's suggestion would be a good place to
start. The task would be to modify RelEx to make use of the link
parser's new
more...
Linas Vepstas <linasvepstas@gmail.com> Aug 16 02:13PM -0500
^ > Following this, testing against a larger corpus as I suggested would
> be valuable..
Simply finding one that is compatbile or almost-compatible, and determining
just how different it is,
more...
Ben Goertzel <ben@goertzel.org> Aug 16 12:24PM -0700
^ Well what I found was a script for converting the Penn Treebank and
Brown Corpus to dependency-grammar format...
The task then becomes to try to tweak this script to give
Stanford-parser-style
more...
--
You received this message because you are subscribed to the Google Groups "link-grammar" group.
To post to this group, send email to link-grammar@googlegroups.com.
To unsubscribe from this group, send email to link-grammar+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/link-grammar?hl=en.
No comments:
Post a Comment