Summary of the Multimodal Interaction Working Group face to face meeting (October 6-8, 2003)

The W3C Multimodal Interaction Working Group held a face to face
meeting in Budapest, Hungary, October 6-8, 2003, hosted by
ScanSoft. There were 41 attendees from 30 organizations.  This note
summarizes the results of the meeting.

The Working Group's highest current priority is defining how modality
components, such as components for processing speech or ink, are
integrated into a host environment.  This work continues and
elaborates on the ideas defined in Section 5 of the Framework Note [1]
which propose a general approach to component interfaces based on the
DOM [2]. At the face-to-face meeting, we reviewed the ongoing work on
the general framework and also reviewed several examples of how
specific modality components fit into this integration framework. The
group is targeting mid January as a date for publishing a Working
Draft based on this work.

Other group activities were reviewed, including:

1. The ongoing work on EMMA (Extensible MultiModal Annotation)[3]
for representing and annotating user input. An update to the Working
Draft is planned for mid-November.

2. The representation of digital ink [4]. An update to the Working
Draft is planned for mid-January.

3. Work on making system and environment information available to
multimodal applications. Because the system and environment work is
closely related to Device Independence (DI)[5], the group
is working closely with the DI group in this area. The group is targeting
mid-January for a Working Draft based on this work.

4. Approaches to handling composite input; that is, coordinated input
from multiple modalities, such as speech combined with a pointing
gesture.

The group also initiated a study of existing approaches to
interaction management and initiated another discussion on
issues in session management.

The face to face meeting included a joint meeting with the Voice
Browser Working Group [6] in which we discussed issues relating to
speech in multimodal contexts, such as how a speech modality object
relates to the general modality integration framework.

During the joint meeting we also discussed some potential requirements
that the MMI group may feed back to the Voice Browser group with
respect to the Speech Recognition Grammar Specification (SRGS) [7] and
the Semantic Interpretation specification [8].

ATT, Voice Genie, Scansoft, Loquendo, T-Systems, and IBM presented
demonstrations of some of their internal multimodal applications.

The next face to face meeting will take place March, 2004, in Cannes,
France, in conjunction with the W3C Technical Plenary meeting (March
1-5).

References:

[1] Multimodal Interaction Framework Note:
    http://www.w3.org/TR/mmi-framework/
[2] DOM home page: http://www.w3.org/DOM/
[3] EMMA: http://www.w3.org/TR/emma/
[4] InkML: http://www.w3.org/TR/InkML/
[5] Device Independence: http://www.w3.org/2001/di/
[6] Voice Browser: http://www.w3.org/Voice/
[7] SRGS: http://www.w3.org/TR/speech-grammar/
[8] SI: http://www.w3.org/TR/semantic-interpretation/

Best regards,

Debbie Dahl, MMI Working Group Chair

Received on Tuesday, 4 November 2003 09:19:15 UTC