W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 2002

Vowels

From: Lisa Seeman <seeman@netvision.net.il>
Date: Tue, 20 Aug 2002 19:29:28 -0700
To: "Wendy A Chisholm (E-mail)" <wendy@w3.org>, jasonw@ariel.ucs.unimelb.edu.au, GV@TRACE.WISC.EDU
Cc: "W3c-Wai-Gl@W3.Org (E-mail)" <w3c-wai-gl@w3.org>
Message-id: <007c01c248ba$960e9450$7200000a@patirsrv.patir.com>

New bees  (and some of the rest of us) may want to look up the Hebrew vowels
threads to understand this email.

The Techno project, the IISO and a bunch of other Peaple all point to Prof.
Yaakov Choueka of Bar-Ilan university. I feel enough people have recommended
his expertise in this. As far as I know, this problem has been solved with
an accuracy of about 95%, but is being used  by a commercial product called
"NAKDAN".

He is holding on to the software license. I did some more mining and Melingo
licenses his algorithm. I have contacted both Yaakov Choueka  and Melingo.
If I  can get a license for a server solution  then I can set up a free
portal that reformats pages and puts in the vowels for accessibility
purposes.
Let me see how much the license would cost. The  cost  may be prohibitive.

But even at 95% success (Yaakov Choueka estimate  - Melingo claim a bit
higher) that is not perfect. We could ask that people put in vowels in words
were  the free service does not translate it correctly.

In other words the checkpoint as it now stands:
'Provide information needed for unambiguous decoding of the characters and
words in the content' would still hold  -  when the portal (assuming I can
build one ) decipher the word correctly, then the checkpoint has been
fulfilleled automaticly. When the portal adds the vowels _incorctly_, then
clearly not all the  information needed for unambiguous decoding has been
provided, and the responsibilty falls on the web content provier to provide
more information.
In the case of Arrabic, vowels would be required until somone makes the same
type of service.

Do we want to say that it can be unambiguous decoding at an affordable cost
(whatever that means). In other words, if a program can decode the word, but
it costs the end user three mounth salary to do so - the job ain't done?

However the problem has been solved from a different perspective -
reasonability. If this algorithm is anything like it seas it is, then
providing vowels even without a portal, is now a lot more affordable. The
cost for a site owner to use software to put in the vowels and then tweak
it, is nothing like as high or as difficult as adding the vowels by hand.

By the way the Israel Internet Society is truly starting to promote
accessibility and WCAG - so it may yet happen hear. So now among my other
hats I am also working for them (yes, as usual, as a volunteer)


And yes, I will continue reporting back. (Whether anyone is interested or
not)

All the best,

Lisa Seeman

UnBounded Access

Widen the World Web

http://www.UBaccess.com



-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org]On
Behalf Of Jason White
Sent: Wednesday, August 14, 2002 5:27 PM
To: Web Content Guidelines
Subject: Agenda



Thursday, 15 August, 2000 UTC (4 PM US Eastern, 10 PM France, 6 AM
Eastern Australia) on +1-617-761-6200, passcode 9224:

This week we plan to continue with last week's agenda, starting with
items 1 and 2 (i.e., further consideration of levels, followed by
assurance requirements and the remaining items on last week's agenda).
Please be sure to read the relevant mailing list threads prior to the
meeting. Last week's agenda is available at
http://lists.w3.org/Archives/Public/w3c-wai-gl/2002JulSep/0125.html
Received on Tuesday, 20 August 2002 12:29:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:19 GMT