Skip to Content.
Sympa Menu

illinois-ml-nlp-users - Re: [Illinois-ml-nlp-users] Including a gazetteer list in Lbj

illinois-ml-nlp-users AT lists.cs.illinois.edu

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

List archive

Re: [Illinois-ml-nlp-users] Including a gazetteer list in Lbj


Chronological Thread 
  • From: Lev-Arie Ratinov <ratinov2 AT uiuc.edu>
  • To: James Clarke <clarkeje AT gmail.com>, anamasiprodriguez AT gmail.com, Nick Rizzolo <rizzolo AT gmail.com>, Mark Sammons <mssammon AT illinois.edu>, illinois-ml-nlp-users AT cs.uiuc.edu
  • Subject: Re: [Illinois-ml-nlp-users] Including a gazetteer list in Lbj
  • Date: Fri, 10 Sep 2010 09:55:58 -0500
  • List-archive: <http://lists.cs.uiuc.edu/pipermail/illinois-ml-nlp-users>
  • List-id: Support for users of CCG software <illinois-ml-nlp-users.cs.uiuc.edu>

Hi Ana.

Sorry for the late response. If you have a gazetteer for PER, the
dirty trick is to add the lines from that gazetteer to one of the
existing PER lists. For example, Data/KnownLists/known_name.lst or to
Data/KnownLists/WikiPeople.lst (if you're very confident in your
gazetter) or Data/KnownLists/known_names.big.lst if you're less
confident or to Data/KnownLists/WikiPeopleRedirects.lst if you're
really not confident.

The right way to do it is to add the Gazetteer as a separate file to
the KnownLists, as you're doing, but then you need to retrain the
system to learn which entity type (PER/LOC/ORG) to associate it with,
and to which degree the system can trust your list. However, I know
that few people have the CoNLL training data and I cannot distribute
it because it's copyrighted. So I don't know if it's an easy option
for you.

Hope that helps.
Best

On Fri, Sep 10, 2010 at 7:47 AM, James Clarke
<clarkeje AT gmail.com>
wrote:
> Did you see this message?  If you craft a reply I can send it to the list.
>
> Begin forwarded message:
>
> From: Ana
> <anamasiprodriguez AT gmail.com>
> Date: 8 September 2010 06:36:09 EDT
> To:
> illinois-ml-nlp-users AT cs.uiuc.edu
> Subject: [Illinois-ml-nlp-users] Including a gazetteer list in Lbj
>
> Hi!
> I downloaded the Illinois Name Entity Tagger from the web
> (LbjNerTagger1.11.release) and I want to add a gazetteer list to the
> application.
> I included my gazetteer list (.lst extension) to the directory
> /Data/Knownlists/. I would like to add this gazetteer to be detected as a
> "Person" name entity. Now I don't know how to continue....
> When I run Lbj, it detects the gazetteer list when loading gazetteers, but
> it doesn't detect the words in the text.
> I know have to include this gazetteer list in some file or in the java
> implementation, but I dont know where.
> Can somebody help me?
> Thank you in advance,
> Ana
> _______________________________________________
> illinois-ml-nlp-users mailing list
> illinois-ml-nlp-users AT cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/illinois-ml-nlp-users
>
>



--
Peace&Love





Archive powered by MHonArc 2.6.16.

Top of Page