[Corpora-List] Sentence similarity

Philipp Koehn pkoehn at inf.ed.ac.uk
Tue Mar 31 20:09:23 UTC 2009


Hi,

there are a number of MT test sets released by NIST and
available at LDC that contain multiple English sentence
of the same foreign (Chinese/Arabic) sentence. They should
be similar and you can use them as positive examples
(negative examples may be random sentences or distorted
examples from the corpus).

-phi

On Sun, Mar 29, 2009 at 3:35 PM, hamed <h_khanpour at yahoo.com> wrote:
> Dear Corpra members
>
> I have developed a system to measure similarity between sentences. but  I do
> not know how to evaluate it? I'm looking for a sentence-similarity corpus,
> i.e., a collection of
>  sentences with manually assigned similarities to other sentences. Any
> ideas?
>
> Thank you very much.
>
> Hamed Khanpour
>
> Computer science student
>
>
>
>
> _______________________________________________
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora
>
>

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora



More information about the Corpora mailing list