Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[java-nlp-user] Stanford NLP: updated versions of everything

Alexander Yeh asy at mitre.org
Wed May 25 17:31:52 PDT 2011


Thank you for the update.

I tried this on the example given in
http://nlp.stanford.edu/software/corenlp.shtml

"Stanford University is located in California.
It is a great university."

The coreference output for this used to be
(from http://nlp.stanford.edu/software/corenlp_output.html):

<coreference>
  <link>
   <source sentence="2" token="5"/>
   <destination sentence="2" token="1"/>
   </link>
</coreference>

which displays as (from http://nlp.stanford.edu/software/example.xml):

"source (sentence 2, token 5) -- destination (sentence 2, token 1)"

The coreference output for this is now
(from the attached sample-input-frm-Stanford-instructions.txt.xml):

     <coreference>
       <coreference>
         <mention representative="true">
           <sentence>2</sentence>
           <start>3</start>
           <end>6</end>
           <head>5</head>
         </mention>
         <mention>
           <sentence>2</sentence>
           <start>1</start>
           <end>2</end>
           <head>1</head>
         </mention>
       </coreference>
     </coreference>

where the current "head" takes the place of the "token" in the old 
format, and the entire chunk is marked between "start" and "end" (not 
including "end").

Unfortunately, the current style sheet given
(attached as CoreNLP-to-HTML.xsl)
cannot handle this new coreference format and displays nothing in the 
coreference link area.

A current browser display for this example is attached as
current-browser-display-of-example.pdf

Thank you
-Alex

John Bauer wrote:
> Stanford NLP is pleased to announce the following updates:
>
> CoreNLP: coref output updated so it no longer takes quadratic space,
> allows pretokenized input
>
> Tagger: now supports TSV input files
>
> NER: updated to be compatible with the NER found in CoreNLP
>
> Parser: minor bug fixes
>
> Segmenter, tregex, classifier: all updated to avoid .jar conflicts
>
> All of these releases are available at:
> http://nlp.stanford.edu/software/index.shtml
> _______________________________________________
> java-nlp-user mailing list
> java-nlp-user at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/java-nlp-user

-------------- next part --------------
A non-text attachment was scrubbed...
Name: sample-input-frm-Stanford-instructions.txt.xml
Type: text/xml
Size: 8638 bytes
Desc: not available
URL: <http://mailman.stanford.edu/pipermail/java-nlp-user/attachments/20110525/4c8979b1/attachment.xml>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CoreNLP-to-HTML.xsl
Type: text/xml
Size: 3795 bytes
Desc: not available
URL: <http://mailman.stanford.edu/pipermail/java-nlp-user/attachments/20110525/4c8979b1/attachment.xsl>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: current-browser-display-of-example.pdf
Type: application/x-pdf
Size: 32337 bytes
Desc: not available
URL: <http://mailman.stanford.edu/pipermail/java-nlp-user/attachments/20110525/4c8979b1/attachment.pdf>


More information about the java-nlp-user mailing list