[Corpora-List] Looglefight

[Jason Eisner]

On Fri, Oct 1, 2010 at 11:29 AM, Roman Klinger <
roman.klinger at scai.fraunhofer.de> wrote:

>  On 10/01/2010 04:25 PM, Adam Kilgarriff wrote:
>
>> Have you ever been writing an NLP paper and been unsure about the usual
>> turn of phrase for something, in the NLP literature?  Now you can ask -
>> at http://looglefight.com - and be answered!
>>
>
> I am curious:
> Did you analyze if the results differ from googlefight.com?
>

Here is one instance where they differ:

On googlefight, "natural language processing" beats "computational
linguistics" (you need to include the quotes), 186K to 139K.  On
looglefight, "computational linguistics" is  ahead, 10K to 4K.  :-)

You shouldn't take this as indicating the relative value placed on science
and engineering.  The looglefight counts are heavily biased by the fact that
these terms (especially "computational linguistics") show up in the
bibliography entries and author affiliations of many papers, since they
appear in the names of conferences, journals, and institutes.

-jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20101001/a67b95b4/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora