<div dir="ltr">Dear Stephan,<div><br></div><div style>Thanks a million for the reply, I was starting to go crazy on LKB. I will try to do what you suggested, but for now the simplest fix for me is to just use the online version of ERG. I wrote lisp code that sends the sentence in an http post request and would parse the response to extract the output, I just hope the html of the page doesn't change soon :). Thank you again.</div>
<div style><br></div><div style>Regards,</div><div style><br></div><div style>Mostafa Gomaa</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Sat, May 18, 2013 at 11:42 PM, Stephan Oepen <span dir="ltr"><<a href="mailto:oe@ifi.uio.no" target="_blank">oe@ifi.uio.no</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">hi mostafa,<br><br>the on-line demonstrator of the ERG is built around PET<br>for parsing, where there are mechanisms for unknown<br>
word handling and parse ranking unavailable in the LKB.<br>furthermore, the LKB by default does not apply packing<br>
during parsing and (on top of that) is about a factor of 5<br>at least slower than PET.<br><br>thus, markedly lower coverage could be owed to unknown<br>words or resource exhaustion, or both.<br><br>given the examples you suggest, maybe you can add the<br>
missing vocabulary to ‘lexicon.tdl’. for packing and parse<br>ranking, these can in principle be enabled. please have a<br>look in ‘parse.lsp’ and ‘active.lsp’; if you need help, i can<br>refresh my memory of that code sometime next week …<br>
<br>best, oe<br><br>nb: there is a lisp API to PET, but there is no way it could<br>be made to work on windoze easily, i fear.<div class="HOEnZb"><div class="h5"><br><br><br>On Friday, May 17, 2013, Mostafa Gomaa <<a href="mailto:mostafa.gomaa89@gmail.com" target="_blank">mostafa.gomaa89@gmail.com</a>> wrote:<br>
> Good day,<br>> I am sorry if this has been asked before, but I am having some problems with LKB+ERG. I am using the latest version of both LKB and ERG, and there are a lot of sentences where LKB doesn't return any parses. I have played around with the online version of ERG and those sentences return parses. Any ideas on what to do about this?. Also I am primarily looking for a grammar with reasonable coverage of English, but doesn't have to be necessarily too large like the ERG. I basically want to be able to use sentences such as:<br>
> "fido is a dog"<br>> "all dogs bark"<br>> "what does fido do?"<br>> "does fido bark?"<br>> "is fido a dog?".<br>> If anyone already knows from where I can obtain such a grammar or how to fix the issue with the ERG, I would be extremely grateful. Also, as a suggestion, I think it would be a very good idea to make an API for the online version of ERG, I think it would be extremely useful and much easier to use. Thank you all in advance.<br>
> Regards,<br>> Mostafa Gomaa<br>><br>><br><br></div></div><span class="HOEnZb"><font color="#888888">-- <br>+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++<br>+++ Universitetet i Oslo (IFI); Boks 1080 Blindern; 0316 Oslo; (+47) 2284 0125<br>
+++ --- <a href="mailto:oe@ifi.uio.no" target="_blank">oe@ifi.uio.no</a>; <a href="mailto:stephan@oepen.net" target="_blank">stephan@oepen.net</a>; <a href="http://www.emmtee.net/oe/" target="_blank">http://www.emmtee.net/oe/</a> ---<br>
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++<br><br>
</font></span></blockquote></div><br></div>