[Corpora-List] arabic transliteration

[Hardie, Andrew]

Hi all

Since lots of people have been mentioning the Buckwalter transliteration, it's worth mentioning that this is designed for a very specific purpose: one-to-one lossless representation of the original sequence of Arabic characters using only ASCII characters. This is ideal for the places where that transliteration is usually used, i.e. within Arabic processing systems. Given the NLP interests of many list members it's not surprising the Buckwalter system was the first suggested.

But this scheme is not ideal for purposes other than automatic processing. It's not designed for readability, or to reflect pronunciation, or to be used (for instance) for the rendering of Arabic examples in published papers! For these kinds of purposes, you are much better off with either an actual IPA transcription (if you need to discuss phonetic details) or an accepted standard Romanisation, such as ISO-233 or the Library of Congress scheme. Typically a Romanisation scheme will not be a direct transliteration because it will normally make explicit features of the pronunciation which are implicit in Arabic.

best

Andrew.

From: corpora-bounces at uib.no [mailto:corpora-bounces at uib.no] On Behalf Of Marwa GRAJA
Sent: 29 May 2012 13:02
To: Amira Barhoumi
Cc: Corpora at uib.no
Subject: Re: [Corpora-List] arabic transliteration

Hi Amira,

You can find in the attached file the java program of Backwalter Code .

Good luck
2012/5/25 Amira Barhoumi <amirabarhoumi29 at yahoo.fr<mailto:amirabarhoumi29 at yahoo.fr>>
Hello,

I want to know how doing transliteration of an Arabic sentence.
Is there any norm?

Thanks,
Amira
Master student in computer science




_______________________________________________
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



--
*****************************************************
Marwa GRAJA
PHD Student
https://sites.google.com/site/marwagraja/

MIRACL Laboratory
www.miracl.rnu.tn<http://www.miracl.rnu.tn>

Faculty of Economic Sciences and management of Sfax
ANLP Research Group
http://sites.google.com/site/anlprg
*****************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20120529/1ef83791/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