[Corpora-List] Free Paraphrases of Noun Compounds -- a task at *SEM 2013

Diarmuid Ó Séaghdha do242
Tue Jan 22 21:10:50 CET 2013


"Onion tears" certainly is attested:

Onion tears affect my contact lenses (http://www.dailymail.co.uk/femail/article-82510/Onion-tears-affect-contact-lenses.html) No more onion tears! (http://durhamfoodie.blogspot.co.uk/2007/12/no-more-onion-tears.html) I actually never have a problem because I wear contacts and for some reason there's barrier between me and those dang onion tears. (http://joannagoddard.blogspot.co.uk/2010/12/chopping-onions.html) I echo the previous commenter that contacts are the best solution for onion tears. (http://www.thekitchn.com/tip-let-cut-onions-go-for-a-sw-117149)

My intuition is that it's a semantically straightforward compound (assuming you know about chopping onions).

There's a debate to be had about whether or not all possible N+N compounds are meaningful to some degree, but for the SemEval task we're not using any unattested compounds.

Diarmuid

On 22/01/13 19:50, E. Bashir wrote:
> I notice (from the website) that there is no option like "not a
> felicitous noun compound" or "not a meaningful noun compound" listed
> among the choices. For example, the noun compound suggested "onion
> tears" is not a noun compound that I have ever heard used, or would ever
> use.
>
> Elena Bashir
>
> ------------------------------------------------------------------------
> *From:* Lushan Han <lushan1 at umbc.edu>
> *To:* Stan Szpakowicz <szpak at eecs.uottawa.ca>
> *Cc:* corpora at uib.no
> *Sent:* Tuesday, January 22, 2013 12:22 PM
> *Subject:* Re: [Corpora-List] Free Paraphrases of Noun Compounds --
> a task at *SEM 2013
>
> Would it be more interesting if some paraphrases are intentionally
> made "wrong"? IMHO, it seems that ranking all true paraphrases is
> not very interesting.
>
> Best regards,
>
> Lushan
>
> On Mon, Jan 21, 2013 at 11:20 AM, Stan Szpakowicz
> <szpak at eecs.uottawa.ca <mailto:szpak at eecs.uottawa.ca>> wrote:
>
> Free Paraphrases of Noun Compounds -- a task at *SEM 2013
>
> We invite all and sundry to participate. All details appear on
> the official site at:
>
> http://www.cs.york.ac.uk/semeval-2013/task4/
>
> Registration for task participants:
>
> http://www.cs.york.ac.uk/semeval-2013/index.php?id=registration
>
> Contact: iris at clul.ul.pt <http://clul.ul.pt/>
>
> Just to whet your appetite, here is the abstract:
>
> The task of classifying English noun compounds explores the idea
> of interpreting the semantics of noun compounds via paraphrases.
> Given a two-word noun compound, the participants are asked to
> produce an explicitly ranked list of its free paraphrases. The
> list will be automatically compared and evaluated against a
> similarly ranked list of paraphrases proposed by human
> annotators, recruited and managed through Amazon's Mechanical
> Turk. The comparison of raw paraphrases will be sensitive to
> syntactic and morphological variations. The ranking of
> paraphrases will be based on their relative popularity among
> annotators. To achieve a reliable ranking, highly similar
> paraphrases will be grouped so as to downplay superficial
> differences in syntax and morphology.
>
> And the relevant dates (slightly tentative), all in 2013:
>
> February 15: Test set ready
> February 15: Registration deadline for task participants
> March 1-15: Evaluation period
> April 9: Paper submission deadline
> May 4: Camera-ready paper due
> June 13 or 14: Workshop co-located with NAACL
>
> Cristina, Iris, Su, Zornitsa, Preslav, Diarmuid, Stan and Tony
>
> --
> Stan Szpakowicz, PhD, Professor
> EECS, Computer Science, University of Ottawa
>
> _________________________________________________
> UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
> Corpora mailing list
> Corpora at uib.no <mailto:Corpora at uib.no>
> http://mailman.uib.no/listinfo/corpora
>
>
>
> _______________________________________________
> UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
> Corpora mailing list
> Corpora at uib.no <mailto:Corpora at uib.no>
> http://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
>



More information about the Corpora mailing list