W3C home > Mailing lists > Public > public-xg-htmlspeech@w3.org > January 2011

RE: questionnaire results and my recommendations (v2)

From: Young, Milan <Milan.Young@nuance.com>
Date: Thu, 27 Jan 2011 11:21:58 -0800
Message-ID: <1AA381D92997964F898DF2A3AA4FF9AD0A075A8C@SUN-EXCH01.nuance.com>
To: "Dan Burnett" <dburnett@voxeo.com>
Cc: "Michael Bodell" <mbodell@microsoft.com>, <public-xg-htmlspeech@w3.org>
Sorry Dan, I really didn't mean to suggest that you were trying to
"screw any one over".  I was just trying to figure out how these
categorizations would affect proposals.

 

What do you think about publishing a simple % to each of the
requirements instead of low/medium/high ratings?  That way there are no
connotations.

 

Thanks

 

________________________________

From: Dan Burnett [mailto:dburnett@voxeo.com] 
Sent: Thursday, January 27, 2011 10:46 AM
To: Young, Milan
Cc: Michael Bodell; public-xg-htmlspeech@w3.org
Subject: Re: questionnaire results and my recommendations (v2)

 

Milan, please re-read my actual text.  I believe I said "Proposals that
support more of the "moderate interest" requirements are more likely to
gain consensus.  Thus, it would be wise for proposals to support as many
of these as possible."  Please note that at no point did I say that
proposals should not or must not address any of the requirements.  I did
also say that we can change the category names if there is concern with
them; your email is the first to suggest possible concern.

 

I can tell you from experience that 50-70% interest in no way guarantees
that the group as a whole will reach consensus on an item, although
proposals that do not support such an item are not likely to gain
consensus.  And consensus is what W3C and broadly-supported standards
are about.

 

Along those lines, there is no plan to "vote" on the proposals.  W3C is
a consensus-based organization, where the goal is to work together to
produce something on which we can have consensus.  If you'll notice, I
did not call the recent questionnaire a "vote", or require majority
decision, or any such thing.  Instead, I very carefully described what
we as group members can reasonably expect of our final proposals based
upon the input from the prioritization survey.

 

Now, how are we going to reach consensus?  I am not describing how in
advance.  That is not because I have a secret agenda on how to screw
everyone over.  Rather, because I have been doing this for a very long
time, I watch how the group operates, notice where we tend to have
agreement and where not, and then suggest ways to proceed that will
allow us, as a group, to reach consensus.  If you have concerns with my
(and/or Michael's) ability to fairly and smoothly assist a group to
reach consensus, please contact us individually.

 

-- dan

 

On Jan 27, 2011, at 12:27 PM, Young, Milan wrote:





I was waiting for today's call to clarify these results.  :-)

 

In particular, I'm unsure about the low/medium/high priorities Dan
assigned.  It seems harsh  to assign a "moderate" rating to features
that 50-70% of us thought would be a necessary part of v1.

 

Perhaps you could clarify how the voting on the proposals will proceed.
Are we going to require a majority decision to approve?

 

Thanks

________________________________

From: public-xg-htmlspeech-request@w3.org
[mailto:public-xg-htmlspeech-request@w3.org] On Behalf Of Michael Bodell
Sent: Thursday, January 27, 2011 7:25 AM
To: Dan Burnett; public-xg-htmlspeech@w3.org
Subject: RE: questionnaire results and my recommendations (v2)

 

It sounds like there is no disagreement with this point so I will update
the requirements document to reflect these results.  Based on our
schedule from the face to face this concludes our prioritized
requirements which was to be done by the end of January (on track).  And
with no one raising email objections about the recommendations this
means we do not need a conference call today to discuss this, so the
conference call is canceled.

 

The next stage of our work is to collect proposals.  We have until the
end of February for this stage.  However, sooner is better as it gives
the group more time to read and digest the proposals and possibly to
discuss them over a call.  I'd suggest that all of the proposals should
evaluate themselves against the prioritized requirements.  For each
requirement I'd suggest categorizing it as one of "The proposal clearly
meets this requirement in a straightforward way", "The proposal can meet
the requirement, but it may not be obvious or straightforward how this
is done", "The proposal meets most of the requirement", "The proposal
meets a little of the requirement", "The proposal doesn't meet any of
the requirement".  That categorization can then help drive some of our
discussions of the propsals.

 

 

________________________________

From: public-xg-htmlspeech-request@w3.org
[public-xg-htmlspeech-request@w3.org] on behalf of Dan Burnett
[dburnett@voxeo.com]
Sent: Tuesday, January 25, 2011 5:38 AM
To: public-xg-htmlspeech@w3.org
Subject: questionnaire results and my recommendations (v2)

Group,

The questionnaire is now closed.  Looking at the results [1] and then
sorting by number of votes I see the following counts:

10 votes for 10 requirements
9 votes for 20 requirements
8 votes for 11 requirements
7 votes for 6 requirements
6 votes for 5 requirements
5 votes for 3 requirements
4 votes for 1 requirement

3 votes for 3 requirements
2 votes for 2 requirements

Based on natural breakpoints and the fact that 5 votes is the halfway
point out of 10, I would suggest 8-10 votes represent "strong interest"
in the requirement, 5-7 votes represent "moderate interest" in the
requirement, and 0-4 votes represent "mild interest" in the requirement.
The requirements are listed in these categories at the end of this
email.


We can discuss and debate names for these different levels the next time
we have a call, but it seems to me that we can at least conclude the
following:

1) Proposals that fail to support the "strong interest" requirements are
unlikely to gain consensus.  Practically, then, as a group we are likely
to require any proposal to support these requirements.

2) Proposals that support more of the "moderate interest" requirements
are more likely to gain consensus.  Thus, it would be wise for proposals
to support as many of these as possible.

3) Gaining consensus to support any of the "mild interest" requirements
will be difficult at best.


I recommend that we add a section to the requirements document that
references the questionnaire results and lists the requirements grouped
into these three different categories.  If there is disagreement on the
names of the categories we can discuss that on a call.

At this point I believe we are ready to consider proposals.  If you
disagree, please send email to the list and we can discuss.









"Strong Interest" Requirements
- FPR40. Web applications must be able to use barge-in (interrupting
audio and TTS output when the user starts speaking).
- FPR4. It should be possible for the web application to get the
recognition results in a standard format such as EMMA.
- FPR24. The web app should be notified when recognition results are
available.
- FPR50. Web applications must not be prevented from integrating input
from multiple modalities.
- FPR59. While capture is happening, there must be a way for the web
application to abort the capture and recognition process.
- FPR52. The web app should be notified when TTS playback finishes.
- FPR60. Web application must be able to programatically abort tts
output.
- FPR38. Web application must be able to specify language of
recognition.
- FPR45. Applications should be able to specify the grammars (or lack
thereof) separately for each recognition.
- FPR1. Web applications must not capture audio without the user's
consent.
- FPR19. User-initiated speech input should be possible.
- FPR21. The web app should be notified that capture starts.
- FPR22. The web app should be notified that speech is considered to
have started for the purposes of recognition.
- FPR23. The web app should be notified that speech is considered to
have ended for the purposes of recognition.
- FPR25. Implementations should be allowed to start processing captured
audio before the capture completes.
- FPR26. The API to do recognition should not introduce unneeded
latency.
- FPR34. Web application must be able to specify domain specific custom
grammars.
- FPR35. Web application must be notified when speech recognition errors
or non-matches occur.
- FPR42. It should be possible for user agents to allow hands-free
speech input.
- FPR48. Web application author must be able to specify a domain
specific statistical language model.
- FPR54. Web apps should be able to customize all aspects of the user
interface for speech recognition, except where such customizations
conflict with security and privacy requirements in this document, or
where they cause other security or privacy problems.
- FPR51. The web app should be notified when TTS playback starts.
- FPR53. The web app should be notified when the audio corresponding to
a TTS <mark> element is played back.
- FPR5. It should be easy for the web appls to get access to the most
common pieces of recognition results such as utterance, confidence, and
nbests.
- FPR39. Web application must be able to be notified when the selected
language is not available.
- FPR13. It should be easy to assign recognition results to a single
input field.
- FPR14. It should not be required to fill an input field every time
there is a recognition result.
- FPR15. It should be possible to use recognition results to multiple
input fields.
- FPR16. User consent should be informed consent.
- FPR18. It must be possible for the user to revoke consent.
- FPR11. If the web apps specify speech services, it should be possible
to specify parameters.
- FPR12. Speech services that can be specified by web apps must include
network speech services.
- FPR2. Implementations must support the XML format of SRGS and must
support SISR.
- FPR27. Speech recognition implementations should be allowed to add
implementation specific information to speech recognition results.
- FPR3. Implementation must support SSML.
- FPR46. Web apps should be able to specify which voice is used for TTS.
- FPR7. Web apps should be able to request speech service different from
default.
- FPR9. If browser refuses to use the web application requested speech
service, it must inform the web app.
- FPR17. While capture is happening, there must be an obvious way for
the user to abort the capture and recognition process.
- FPR37. Web application should be given captured audio access only
after explicit consent from the user.
- FPR49. End users need a clear indication whenever microphone is
listening to the user

"Moderate Interest" Requirements
- FPR33. There should be at least one mandatory-to-support codec that
isn't encumbered with IP issues and has sufficient fidelity & low
bandwidth requirements.
- FPR28. Speech recognition implementations should be allowed to fire
implementation specific events.
- FPR41. It should be easy to extend the standard without affecting
existing speech applications.
- FPR36. User agents must provide a default interface to control speech
recognition.
- FPR44. Recognition without specifying a grammar should be possible.
- FPR61. Aborting the TTS output should be efficient.
- FPR32. Speech services that can be specified by web apps must include
local speech services.
- FPR47. When speech input is used to provide input to a web app, it
should be possible for the user to select alternative input methods.
- FPR56. Web applications must be able to request NL interpretation
based only on text input (no audio sent).
- FPR30. Web applications must be allowed at least one form of
communication with a particular speech service that is supported in all
UAs.
- FPR55. Web application must be able to encrypt communications to
remote speech service.
- FPR58. Web application and speech services must have a means of
binding session information to communications.
- FPR6. Browser must provide default speech resource.
- FPR20. The spec should not unnecessarily restrict the UA's choice in
privacy policy.


"Mild Interest" Requirements
- FPR29. Speech synthesis implementations should be allowed to fire
implementation specific events.
- FPR31. User agents and speech services may agree to use alternate
protocols for communication.
- FPR43. User agents should not be required to allow hands-free speech
input.
- FPR10. If browser uses speech services other than the default one, it
must inform the user which one(s) it is using.
- FPR8. User agent (browser) can refuse to use requested speech service.
- FPR57. Web applications must be able to request recognition based on
previously sent audio.



[1] http://www.w3.org/2002/09/wbs/45260/ReqPri02/results

 
Received on Thursday, 27 January 2011 19:22:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 27 January 2011 19:22:46 GMT