W3C home > Mailing lists > Public > www-math@w3.org > April 2003

RE: Content Markup or Presentation Markup for Audio Rendering of MathML

From: Bernhard Keil <Bernhard.Keil@soft4science.com>
Date: Wed, 23 Apr 2003 18:52:06 +0200
To: "Helder Ferreira" <hfilipe@fe.up.pt>
Cc: <www-math@w3.org>, <dfreitas@fe.up.pt>
Message-ID: <IIEOKCCPKDALHBHODAENKEBJCGAA.Bernhard.Keil@soft4science.com>

>but can i use XSLT stylesheet to transform MathML into plaintext offline?
  yes, of course.

>For instance, someone gives you a document, could be, HTML+MathML, but can
>also be in plaintext+MathML.
  The input of a XSLT transformation has to be XML.
  The ouput can be XML or plain text.
  If you have plaintext+MathML as input source 
  you could create a simple parser that selects the MathML islands 
  and transforms them with XSLT.


>There are diferences. For instance, the audio render has to be more precise
>in describing the formula since the blind person has to ear it and
>interpret. Visually it's more easy since your eyes and brain are trainned to
>recognize the formula faster. Try to dictate to other person a complex
>expression and later on show it to the person, and see what he/she
>understands better and interprets faster.
     
I don't agree in general. You should not underestimate blind people.
I know a very good C++ software developer, he is faster and better
than me and he is blind.
He uses a tool wich does not more than speaking the text where
he points in Visual Studio with the mouse. I think doing programming
in C++ is much more complex than understanding a mathematical
formula. He would not want a tool that explains him C++, since
he nows everything about C++. So I don't think that a blind person 
needs ( or wants) in general
more interpretation or explaination than a seeing perso.


>Maybe i've choosed the wrong words, what i want is an interpretation of a
>MathML formula and convert it to something that can go to a TTS engine to be
>rendered as speech.
  I have understood that

>And both HTML docs are rendered in my IE6+MathPlayer Plugin.
   yes, but the visual rendering should be the same.
   What you see is in both cases the same, the visual  "presentation".

>Why do you doubt that Content Markup
>is reach enough?
>Wasn't it created to struture the meaning of the mathematical expressions?
  MathML Content Markup covers only a small little subset of
  Mathematics , Physics, Medical Mathemathics, ... and s.o.
  It might be good enough to cover shool mathematics.


regards,


Bernhard Keil
mailto:Bernhard.Keil@soft4science.com


-----Original Message-----
From: Helder Ferreira [mailto:hfilipe@fe.up.pt]
Sent: Wednesday, April 23, 2003 6:25 PM
To: Bernhard Keil
Cc: www-math@w3.org; dfreitas@fe.up.pt
Subject: Re: Content Markup or Presentation Markup for Audio Rendering
of MathML


> I don't understand why you want to create a new parser to parse MathML.
> MathML is XML and can therefore be parsed with any XML DOM or SAX parser.
> The standard method to transform a XML instance to other formats
> (like in your case) is to create an XSLT stylesheet.
>

Yes, but can i use XSLT stylesheet to transform MathML into plaintext
offline?
For instance, someone gives you a document, could be, HTML+MathML, but can
also be in plaintext+MathML.

You're a blind person and you use a tool to convert the entire document to
plaint text and some prosody information (output that you can use to render
the text in a TTS engine).

Maybe i've choosed the wrong words, what i want is an interpretation of a
MathML formula and convert it to something that can go to a TTS engine to be
rendered as speech.

> As a vendor of an WYSIWYG authoring appllication
> for the creation of XHTML 1.1 + MathML 2.0 docments,
> my oppinion is, that it makes no sense to support
> Content Markup in a WYSIWYG authoring software.
> If it would support Content Markup, than it can not be WYSIWYG
> at the same time. Because you would have to visualy mark things
> that differ in its meaning but not in its visual layout (presentation).
>

I agree.
However, audio rendering of Presentation Markup is less meaningfull than
Content Markup.
Should this be considered a lack in the accessibility features of MathML?

> You can always transform Content Markup to Presentation Markup.
> Transforming  Presentation Markup to Content Markup is not
> possible in general.

That was my opinion too.

>
> If you would see an audio browser playing the same conceptual role like a
visual browser,
> than its behavior should be analog to the behavior of a visual browser.
>

There are diferences. For instance, the audio render has to be more precise
in describing the formula since the blind person has to ear it and
interpret. Visually it's more easy since your eyes and brain are trainned to
recognize the formula faster. Try to dictate to other person a complex
expression and later on show it to the person, and see what he/she
understands better and interprets faster.

> Every existing visual browser shows only Presentation Markup.
> When you look at the rendering result you can not
> determine whether the source was created as Content or Presentation
> Markup.
>

I have two HTML documents in my desktop created with MathPage from the
MathType.
One uses presentation markup, and the other content markup (i've modified
the source by hand, since MathType only generates presentation markup). And
both HTML docs are rendered in my IE6+MathPlayer Plugin.
This were my first experience in placing MathML in HTML docs.. and worked...
Or have i cheated the result in a way that i'm not knowing?

> But if your target audience are blind people there might be the wish, that
> the role of an audio browser goes beyond the one of an visual browser
> in that way that he interpretes and explains the content and not it's
> visual presentation wich might be meaningless for a blind person.
>
> So I think there might be good reasons to choose a markp language
> that describes the content/meaning wich is not bound to any visual
> presentation.
> But it is the question whether MathMl Content Markup is reach
> enough to fullfill that role in general. I doubt that.
>
>
> Bernhard Keil
> mailto:Bernhard.Keil@soft4science.com
>
> -------------------------------------------------------------
> soft4science             Bernhard Keil
> Nibelungenstr. 4        80639 Munich       Germany
> +49 89 / 95 411 088  http://www.soft4science.com
> +49 173 / 72 53 669  http://www.MathML.net
>

Now you understand were i want to get. Why do you doubt that Content Markup
is reach enough?
Wasn't it created to struture the meaning of the mathematical expressions?
Don't interpret me wrong, i'm only questioning because i want to see if i'm
going on the wrong direction or not.

Anyway, thank you for your opinion.

   Helder Ferreira
   Laboratory for Speech Synthesis, Electroacustics, Signals and
Instrumentation
   Faculty of Engineering University of Porto, Portugal
Received on Wednesday, 23 April 2003 12:52:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:12:54 GMT