[developers] Running LKB in tty mode

Emily M. Bender ebender at uw.edu
Mon May 6 16:18:09 CEST 2013

Dear Mostafa,

That should work, but the lkb *parse-record* isn't going to
be ordered in any helpful way --- there is no sense in which
the first element of that structure should be expected to be
the best parse.

In order to get access to parse ranking (and also to be
able to run the ERG over larger sentences), you'll probably
want to use something like the set-up described here:


You would then have the MRS in textual format, which
you would need to read in for further lisp-based processing,
but that is probably worth it, for the gain of parse selection
and a very much more sophisticated grammar.


On Mon, May 6, 2013 at 2:20 AM, Mostafa Gomaa <mostafa.gomaa89 at gmail.com> wrote:
> So sorry for spamming the mailing list, just wanted to let you now that I
> was able to answer my question regarding receiving mrs output. I also wrote
> the following function to save a few keystrokes
> (defun do-get-mrs-tty (sentence)
>   (progn
>     (do-parse-tty sentence)
>     (mrs::extract-mrs (first lkb::*parse-record*))))
> this does the trick for now, suggestions for better ways are more than
> welcome. Thank you all for your patience.
> Regards,
> Mostafa Gomaa
> On Sun, May 5, 2013 at 11:01 AM, Mostafa Gomaa <mostafa.gomaa89 at gmail.com>
> wrote:
>> Dear All,
>> I was finally able to get the code to run (hopefully I don't face more
>> problems). I was able to successfully load a grammar and get back the parse
>> tree using the function do-parse-tty. Now I'd like to get MRS output, can
>> anyone point out the functions that I'd need for that?. I can't seem to find
>> what i'd need in the functions that are exported from the mrs package. Any
>> guidance would be greatly appreciated. Thanks all for the great help and
>> advice I received so far.
>> Regards,
>> Mostafa Gomaa.

Emily M. Bender
Associate Professor
Department of Linguistics
Check out CLMS on facebook! http://www.facebook.com/uwclma

More information about the developers mailing list