Summary of the June 21-23 Multimodal Interaction Working Group face to face meeting

The W3C Multimodal Interaction Working Group met in Chicago on
Thursday and Friday, June 21-23 hosted by Oracle and Motorola. A joint
session was held on June 21 with the W3C Voice Browser Working Group.

During the joint meeting we discussed how the information provided by
EMMA [1] maps into the information maintained in VoiceXML 2.1 (confidence,
nbest, etc.). We also discussed possible ways to represent speaker
verification and identification results in EMMA.  We saw a multimodal demo
from
Motorola and an SCXML [2] demo from IBM. During the multimodal meeting we
saw another multimodal demo from V-Enable and a demo of the multimodal
architecture [3] by Deborah Dahl.

During the Multimodal meeting, we discussed how SCXML could be used in
the MMI architecture, and saw some examples of how SCXML, XHTML and
VoiceXML could be used together to build multimodal applications. We
had a presentation from ATT on the Florence dialog manager and
discussed how it could fit into the Multimodal Architecture. We
discussed several approaches to the data model, but this topic will
need further discussion. For the next Working Draft of the MMI
Architecture, we plan additional discussions on startup, requirements
for XHTML, remote events, and security as well as the data model.

The EMMA implementation report work is making good progress. This is
the major remaining task in EMMA before it transitions to Candidate
Recommendation.

Thanks to Oracle and Motorola for hosting the meeting and related
activities. The next MMIWG face-to-face meeting is scheduled for Oslo,
Norway, September 18-20, 2006, hosted by Opera Software.

best regards,

Deborah Dahl
Chair, Multimodal Interaction Working Group

[1] EMMA: http://www.w3.org/TR/emma/
[2] SCXML: http://www.w3.org/TR/2006/WD-scxml-20060124/
[3] MMI Architecture: http://www.w3.org/TR/2006/WD-mmi-arch-20060414/

Received on Wednesday, 23 August 2006 13:21:02 UTC