[Corpora-List] Word similarity from large text corpus

David Jurgens david.jurgens at gmail.com
Fri May 13 06:28:56 UTC 2011


Hi Pham,

  If you want to use LSA, there's the S-Space
Package<http://code.google.com/p/airhead-research/>,
which has a Java LSA implementation as well as several other co-occurrence
based algorithms in case you need to compare their similarity values.

  Thanks,
  David

On Thu, May 12, 2011 at 11:04 PM, Thomas Meyer <Thomas.Meyer at idiap.ch>wrote:

>  Hi Pham,
>
> If you use WordNet (http://wordnet.princeton.edu/) you can pass your word
> pairs to the following Perl-Module to compute their similarity (implements
> different measures, e.g. Resnik, Lin etc.)
>
>
> http://search.cpan.org/~tpederse/WordNet-Similarity-2.05/lib/WordNet/Similarity.pm
>
> here is a Java class doing the same:
>
> http://nlp.shef.ac.uk/result/software.html
>
> Hope it helps,
> Thomas
>
>
> On 13/05/11 04:53, Minh Pham wrote:
>
> Dear all,
>
> I am looking for a tool which can compute similarity of two words based on
> their co-occurrences in a large text corpus (using LSA or something like
> that).
> Could you please suggest me some tools for that task?
>
> Thank you,
>
> Best regards,
> Pham
>
> --
> Pham Quang Nhat Minh (Mr)
> PhD student
> NLP Laboratory - School of Information Science - JAIST
> 1-1 Asahidai, Nomi, 923-1292 Japan
> Email: minhpqn at jaist.ac.jp
> Web: http://www.jaist.ac.jp/index-e.html
> Phone: (+81) 090-9440-1556
>
>
> _______________________________________________
> UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
> Corpora mailing listCorpora at uib.nohttp://mailman.uib.no/listinfo/corpora
>
>
> _______________________________________________
> UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20110512/db545b93/attachment.htm>
-------------- next part --------------
_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora


More information about the Corpora mailing list