kovvalsky / langpro Goto Github PK

A long awaiting addition that should drastically decrease tableau size and proof runtime.
A good starting example is SICK-train 200.

First command from readme has error: source sink 'my_word_net' does not exist from knowledge.pl

And ccg_sen_d/SICK_trial_sen.pl does not exist.

Perhaps you forgot to commit them ?

Hi! Thank you for making public the codebase for your nice work!
I want to try running it and see the results with different parse inputs,
but I I can't see what are the difference among files in ccg_sen_d.
I thought SICK_trial_ccg.pl and SICK_trial_reccg.pl may be the same
(as these seem parse results of "rebanking" models,)
but through diff command these are different.
Could you add some info on the naming convention for these files??

Thank you in advance!

Hi,

I used esslli_langpro.py to go ahead and run LangPro on the SICK and FraCaS datasets. Very nice, and very useful tool!

However, I was curious if there was an API for individual (and batch) processing, such as I assume is working for the demo here:

https://naturallogic.pro/LangPro/

I wasn't sure where to find code if available for the demo.

Also, it would be convenient if it had a Pengines server.

Before I set about trying to write this API, I thought I would ask, since the documentation appears not to have covered it.

Thank you,
Andrew

I am trying to run the code on datasets other than SICK and FraCaS, however the documentation is scarce regarding to process of converting a dataset to the necessary format (eg. how do you get the files "SICK_trial_ccg.pl" and "SICK_trial_sen.pl" from "SICK_trial.txt").

Additional documentation is needed for the steps necessary to produce these files for any dataset.