RE: [1.1] Exception for text equivalent (non-urgent)

> We work a lot for organizations that handle historical 
> information. In this domain, more and more old documents get 
> digitized. Some of the archives have expressed the wish to 
> publish the digital documents or photographs on their 
> websites, and ask the visitors to help them to make 
> transcriptions of the documents (or describe the content of 
> the photographs). 
> 
> Until the visitors have created a transcription or 
> description, no text equivalent is available. 

Off the list people have asked me why the organizations does not use OCR for
this task. Because of the bad legibility, this is not possible. For a real
world example look at http://www.dutchgenealogy.nl/test/document.jpg. This
is a 16th century document which 99% of the people in the Netherlands would
have trouble reading. Only obsessed genealogists like me can. I do not
believe any OCR software will be able to transcribe this in the next 10
years. Especially since every clerk had their own handwriting and different
areas in the Netherlands had different ways to write the letters. 

In tonight's telecon, a remark was made that a site which puts up these
images without alt isn't accessible. I agree, it isn't accessible. One might
argue that the documents are inaccessible to sighted people too, since 99%
can't read what it says. So that's a level playing field :-) But what could
the archive do to make these documents accessible at level 1 that isn't an
undue burden? They rather spend their budget on digitalization than on
transcriptions, because that's something the audience can help with. Shall
we just forbid them to publish these documents if they want to meet level 1?
That's against everything level 1 should be about. We don't even want to
tell them they can't use black text on a black background at level one,
after all. 

I hope to discuss this at a more appropriate time, when we are not so busy
with the upcoming TR.

Yvette Hoitink
CEO Heritas, Enschede, The Netherlands
E-mail: y.p.hoitink@heritas.nl

Received on Thursday, 4 March 2004 19:02:35 UTC