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

RE: Query about support for Hebrew

From: Lisa Seeman <seeman@netvision.net.il>
Date: Wed, 18 Sep 2002 02:49:51 -0700
To: gv@trace.wisc.edu, w3c-wai-gl@w3.org
Cc: judy@w3.org
Message-id: <006501c25ef8$bec24be0$7200000a@patirsrv.patir.com>
MessageThis is the company that I reported about in my email entitled NAKDAN

Note: they do not have a typical screen reader -  the user has to copy and
paste bits of the web site into their screen reader. The screen reader will
then read the section to you or add vowels. It does not render content as
"interactive" so it can not be used for surfing but for rendering specific
sections of content. You must have partial vision to operate their screen
reader.

I spoke with them a month ago and they had  no concrete plans to develop an
interactive screen reader that could be used by people without vision. It is
a good tool for some learning disabilities, but it is it is priced at about
the average monthly salary. ( most people with learning disabilities earn
less)

The tool for author to put on a specific site, that reads the site to a user
($2950) also only helps users with vision and is not interactive.
The summary of their , and other products is at
http://ubaccess.com/hebrew-access.html

I recommend that we coordinate when talking to these companies. It will give
them a clearer idea of what they need to do to build a screen reader

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 Gregg Vanderheiden
  Sent: Tuesday, September 17, 2002 10:15 PM
  To: w3c-wai-gl@w3.org
  Subject: FW: Query about support for Hebrew


  FYI

  Gregg





  Dear Gregg,

  Further to Judy's reference to our company, Melingo, which is a subsidiary
of Britannica Israel, and the leader in Natural Language Processing for
semitic languages & computational linguistics, has developed a Hebrew TTS
engine - Kolan. Kolan is able to read hebrew texts with or without vowels at
an extremely high degree of accuracy (average accuracy is over 98%).



  The problem of hebrew is quite complex, and together with Arabic poses the
greatest challenge for speech synthesis (as well as for many other
applications such as data search & retrieval). Kolan overcomes the problems
of morphological complexity, lack of vowels, and ambiguity, by incorporating
a context sensitive morphological analyzer, together with a full inventory
(a dictionary) of the hebrew language, as well as a soundex system that
deals with names and non-hebrew words. Of course all this is transparent to
the user of the applications that we have created for Kolan.



  As to applications, Kolan is incorporated into several, such as a screen
reader for people with learning disabilities or impaired vision (sold for
$826), or as a web site owner's tool that enables within minutes the
creation of an accurate reading of the entire site, that can then be
activated by the client (sold for $2950). Kolan is already incorporated also
into telecom applications especially in the field of unified messaging. All
of the above applications are already commercially available, and of course
additional applications can be and are being developed.



  I would be happy to be of further assistance, should you so require.



  Leon Paull

  Product Manager

  Melingo Ltd

  Tel: +972-3-6070423, cell: +972-54-646201

  Fax: +972-3-6070401

  www.melingo.com

  www.ravmilim.co.il

  www.morfix.co.il

  Melingo is a subsidiary of Britannica.com Israel
Received on Wednesday, 18 September 2002 04:52:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:20 GMT