Support for users of CCG software closed 7-27-20

[Dain Kaplan <dk453 AT cam.ac.uk>]

Hi Mark,

Looking forward to it. In the meantime, please find attached the diff for optionally outputting sentence breaks.

DK

Attachment: DocBase-outputnl.diff
Description: Binary data

On Feb 17, 2013, at 4:15 PM, "Sammons, Mark" <mssammon AT illinois.edu> wrote:

Hi, Dain.

I'm glad you like the Coreference tool. Our new version -- which should be out in a few weeks -- will be more flexible with regard to pre-tokenized input.  But if you would like to send the updated code, I'd like to have it so that we can incorporate it into the older coref system at some point.

Best regards,

Mark

Mark Sammons
Principal Research Scientist
Cognitive Computation Group
Department of Computer Science 
University of Illinois at Urbana-Champaign
217-265-6759

---

From: illinois-ml-nlp-users-bounces AT cs.uiuc.edu [illinois-ml-nlp-users-bounces AT cs.uiuc.edu] on behalf of Dain Kaplan [dk453 AT cam.ac.uk]
Sent: Sunday, February 17, 2013 10:10 AM
To: illinois-ml-nlp-users AT cs.uiuc.edu
Subject: [Illinois-ml-nlp-users] LBJCoref and sentence splitting/newlines

Hello,

Nice work on the LBJ coreference annotator. It was very easy to get working. I did have the need, however, to preserve sentences exactly as they were in the input, and output them again (separated by newlines) with the annotations. To do the latter, I modified DocBase#toAnnotationString to optionally output newlines between sentences. Though the changes for this were minimal, I am happy to provide a patch for it. For the former, I had to hack out the use of the SentenceSplitter entirely, but I was wondering if there was a better way to do this? It would be nice if there were a command line flag for toggling "use input as is" like Stanford Core NLP has.

Cheers,
DK