Re: arabic corpus wanted!

[Andrew Freeman <andyf@umich.edu>]

Hello Sonia Halimi,
  Watt's concordance tool works better than any other I am aware of for
looking at corpora in Arabic script.  It  supports Arabic, or more
precisely it doesn't do anything to prevent you from taking advatage of
your system's Arabic capabilities if there is Arabic enabled on your
computer, so if you know how how to take advantage of the Arabic system
on your windows machine, you can get it to work for Arabic.  You can
visually examine a big context around words, although the left right
versus front back can be confusing, since left is assumed to "before."  It
will also generate statistics on frequencies and stuff.
  A big problem that it won't solve for you is lemmatization.  ie albyt
and byt get counted as separate items unless you list all items in that
category.  There is probably a fiendishly clever way around that in
Watt's concordancer but I have opted for 'c' code to solve that problem.

 My advice is to avoid Athelstan's MonoConc Pro.

  You can get Rob Watt's concordance at http://www.rjcw.freeserve.co.uk/

  As for corpora, there is all kinds of stuff on the web, including the
Bible, the Quraan and newspapers.

have fun,
cheers,
andy

 On Thu, 21 Feb 2002 halimi3@etu.unige.ch wrote:

> I would be grateful if you could provide me  with an arabic corpus for my research.
>
> Do you know of a programme which extracts word combinations from arabic texts?
>
> I thank you very much for your cooperation.
>
> Best regards
>
> Sonia HALIMI
> Student in DEA at the University of Geneva
>


********************************************************************
* Andrew T. Freeman                                                *
* Ph.D. Candidate, Arabic Linguistics, Dept. of Near East. Studies *
* U of M, Ann Arbor, Michigan, USA                                 *
* http://www.umich.edu/~andyf                                      *
********************************************************************