[Corpora-List] Metaphor identification

[Rayson, Paul]

Hi,

Thanks to Gill for the mention of our work at Lancaster. For some more details on our method, you should visit the Metaphor in end of life care project website at http://ucrel.lancs.ac.uk/melc/index.php and specifically the following pages with references to prior work and slides from our talks and workshops:

http://ucrel.lancs.ac.uk/melc/Publications.php
http://ucrel.lancs.ac.uk/melc/aca_dissemination.php
http://ucrel.lancs.ac.uk/melc/workshop_jan2014.php

Regards,
Paul.

Dr. Paul Rayson
Director of UCREL and Senior Lecturer in Computer Science
Faculty of Science and Technology Director of International Teaching Partnerships
School of Computing and Communications, InfoLab21, Lancaster University, Lancaster, LA1 4WA, UK.
Web: http://www.comp.lancs.ac.uk/~paul/
Tel: +44 1524 510357 Fax: +44 1524 510492

From: corpora-bounces at uib.no [mailto:corpora-bounces at uib.no] On Behalf Of Gill Philip
Sent: 16 May 2014 17:53
To: Asim Rai
Cc: Corpora at uib.no
Subject: Re: [Corpora-List] Metaphor identification

Dear Muhammed,
the availability of software to aid metaphor extraction is language-dependent, with English being the best-provisioned.  And of course the success or usefulness of any system depends on which definition of metaphor you are using, e.g. X is Y metaphors; vehicle terms (c.f. Cameron 2010); MIP (Pragglejaz 2007) or MIPVU (Steen et al 2010), or something else. So first of all, decide what you want to class as "metaphor", and only then start looking for software that can help you locate it.

There are several options available, but none are fully automatic - they will let you extract preliminary data but you will have to refine it by hand. If your data is in English (or one of the other supported languages) you can try using WMatrix  (ucrel.lacs.ac.uk/<http://ucrel.lacs.ac.uk/>) to identify contrasting semantic domains - this is a popular option for metaphor scholars, and seems to come up with reliable and convincing results.
You might also have a look at the metaphor-tagged BNC-baby to see if the way it was constructed can be of use for your purposes. http://www2.let.vu.nl/oz/metaphorlab/metcor/documentation/home.html  ( it adheres to the MIP / MIPVU view that metaphor is anything that isn't the "central or basic" meaning of a word, though, which means it counts prepositions as metaphors too. You may or may not find that useful for your purposes)

If the language of your corpus is not English, you may find things a bit more difficult. I had some success with italian data, extracting key-word listings and grouping them semantically (by hand) then comparing content-word groupings to the key words (Philip 2010, 2012). This procedure can only work if your corpus is domain-specific, i.e. if the topics it contains are restricted in scope, because only in that was is it possible to identify non-topic related vocabulary as potentially metaphorical.
As a last resort, you could always read your corpus, manually tag it, then work from the tags. But I guess you're trying to avoid that option.
hope this helps,
Gill
Gill Philip. 2010. Metaphorical keyness in specialised corpora. In: Marina Bondi and Mike Scott (eds.), Keyness in Texts .<https://benjamins.com/#catalog/books/scl.41> (pp. 185–204)
Gill Philip. 2012. Locating metaphor candidates in specialized corpora using raw frequency and keyword lists. In: Fiona MacArthur, José Luis Oncins-Martínez, Manuel Sánchez-García and Ana María Piquer-Píriz (eds.), Metaphor in Use: Context, culture, and communication. (pp. 85–106)


On 16 May 2014 18:18, Asim Rai <masimrai at gmail.com<mailto:masimrai at gmail.com>> wrote:
Dear Fellows
I am Dr Asim form Pakistan.
I am currently working on a research where I need to identify the metaphors in my corpus of approximately 2.5 million words.
Is there any software which may help in Metaphor identification

Any help would be appreciated.

Best
Dr Muhammad Asim Mahmood

_______________________________________________
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



--
*********************************
Dr. Gill Philip
Università degli Studi di Macerata
Dipartimento di Scienze della Formazione, dei Beni Culturali, e del Turismo
Piazzale L. Bertelli
Contrada Vallebona
62100 Macerata
Italy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20140518/34953ecd/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