universal grammar in XHTML+Voice

Hi,
I scrawled all posts in this newsgroup but couldn't find answer to my 
question :
I would like to define some specific words that would perform some 
specific actions (like a page change), at anytime in my dialog. I found 
some useful docs 
(http://cafe.bevocal.com/docs/vxml/gettingstarted.html#200141) 
describing how to use the <link> tag, with the grammar
"universal" 
attribute, but my browser (Opera) wouldn't accept these elements.

Is there a way to perform such a Word Spotting, separately from my dialog???
(Basically, just like when you say "Opera (something)" and it performs
a 
  specific action...)
If not, how can I manage to assign a specific action to an element when 
it is spoken??


Thanks in advance for your answer.

Le Tue, 21 Nov 2006 17:09:37 +0200, A.touchot <augustin@kom.aau.dk> a  
écrit:

>
> Hi,
> I scrawled all posts in this newsgroup but couldn't find answer to my  
> question :
> I would like to define some specific words that would perform some  
> specific actions (like a page change), at anytime in my dialog. I found  
> some useful docs  
> (http://cafe.bevocal.com/docs/vxml/gettingstarted.html#200141)  
> describing how to use the <link> tag, with the grammar
"universal"  
> attribute, but my browser (Opera) wouldn't accept these elements.
>
> Is there a way to perform such a Word Spotting, separately from my  
> dialog???
> (Basically, just like when you say "Opera (something)" and it
performs a  
>   specific action...)
> If not, how can I manage to assign a specific action to an element when  
> it is spoken??
>
>
> Thanks in advance for your answer.
>
>
> Augustin.

Hello!

With all due respect to the efforts in various companies to provide  
documentation (such as BeVocal Café), I have to add that they do introduce  
proprietary features and behaviour.

Therefore, I recommend you to always check what you read in the VoiceXML  
specification.

Regarding your specific question: grammars can be scoped. See:

http://www.w3.org/TR/voicexml20/#dml3.1
http://www.w3.org/TR/voicexml20/#dml3.1.3

Hopefully this solves your problem. I'll answer try to answer any further  
questions as well.

Good luck!


-- 
ROBO Design - We bring you the future

Hi,

Thanks for your reply,

    but actually I was talking about XHTML+Voice... and the 'scope' 
attribute with the 'document' value seems to be only available with 
VoiceXML! (http://www.voicexml.org/specs/multimodal/x+v/12/#vxml-scope).

Do you have any idea how to make a word spotting work (to use just like 
in this example http://www.voicexml.org/specs/multimodal/x+v/12/#N10266, 
with some link actions, for a menu), with an existing dialog form ... ??

Thanks in advance for your answer.


Augustin.




Mihai Sucan a écrit :
> Le Tue, 21 Nov 2006 17:09:37 +0200, A.touchot <augustin@kom.aau.dk> a
 
> écrit:
> 
>>
>> Hi,
>> I scrawled all posts in this newsgroup but couldn't find answer to my 

>> question :
>> I would like to define some specific words that would perform some  
>> specific actions (like a page change), at anytime in my dialog. I 
>> found  some useful docs  
>> (http://cafe.bevocal.com/docs/vxml/gettingstarted.html#200141)  
>> describing how to use the <link> tag, with the grammar
"universal"  
>> attribute, but my browser (Opera) wouldn't accept these elements.
>>
>> Is there a way to perform such a Word Spotting, separately from my  
>> dialog???
>> (Basically, just like when you say "Opera (something)" and it
performs 
>> a    specific action...)
>> If not, how can I manage to assign a specific action to an element 
>> when  it is spoken??
>>
>>
>> Thanks in advance for your answer.
>>
>>
>> Augustin.
> 
> 
> Hello!
> 
> With all due respect to the efforts in various companies to provide  
> documentation (such as BeVocal Café), I have to add that they do 
> introduce  proprietary features and behaviour.
> 
> Therefore, I recommend you to always check what you read in the 
> VoiceXML  specification.
> 
> Regarding your specific question: grammars can be scoped. See:
> 
> http://www.w3.org/TR/voicexml20/#dml3.1
> http://www.w3.org/TR/voicexml20/#dml3.1.3
> 
> Hopefully this solves your problem. I'll answer try to answer any 
> further  questions as well.
> 
> Good luck!
> 

Le Wed, 22 Nov 2006 20:02:51 +0200, A.touchot <augustin@kom.aau.dk> a  
écrit:

> Hi,
>
> Thanks for your reply,
>
>     but actually I was talking about XHTML+Voice... and the 'scope'  
> attribute with the 'document' value seems to be only available with  
> VoiceXML! (http://www.voicexml.org/specs/multimodal/x+v/12/#vxml-scope).
>
> Do you have any idea how to make a word spotting work (to use just like  
> in this example http://www.voicexml.org/specs/multimodal/x+v/12/#N10266,  
> with some link actions, for a menu), with an existing dialog form ... ??
>
> Thanks in advance for your answer.
>
>
> Augustin.

Hello!

That's a mistake I've made. I was in a hurry and I didn't check the X+V  
spec as well.

You are right, but as I knew, it's possible to have a grammar per form and  
another grammar per item.

Since I have never needed this functionality I have asked Ola P. Kleiven  
 from Opera Software for clarifications. He said the "grammar can be placed
 
both as a child of a form and as a child of a field, if the field grammar  
does not match, the form grammar wil be searched". This provides you with 

a similar effect.

The link you have provided  
( http://www.voicexml.org/specs/multimodal/x+v/12/#N10266 ) describes a  
functionality which is supported in Opera. You can have such a grammar as  
a child of a form.

Let me know if this works for you.

Good luck!

-- 
ROBO Design - We bring you the future