Rationale for the Emotion Incubator group

Dear participants to the Emotion Incubator group,

now that most of us are back from holidays, it is time to get up to 
speed with the new W3C group on Emotions. Let me start up the discussion 
with my view of the group's rationale. Your views and discussions on 
this are very welcome!

In the HUMAINE network [1], which deals with emotion-oriented computing 
in a very broad sense, we have spent 1 1/2 years discussing and 
proposing an XML language for representing and annotating emotions in 
various technological contexts. The result of this discussion is the 
HUMAINE Emotion Annotation and Representation Language (EARL) [2]. It 
attempts to describe emotions in a very generic sense, i.e. not bound to 
specific applications nor to specific sets of emotion category labels -- 
not even to using category labels at all! Instead, we have made concrete 
proposals on a mechanism for using custom sets of labels or of 
continuous scales motivated by emotion theory, and have outlined the 
idea of using the emotion language as a "plug-in" language in 
combination with task-specific languages such as EMMA [3], SSML [4], or 
others.

The W3C Emotion Incubator group is a move to open up the discussion to a 
wider audience, to critically assess the suggestions made by the HUMAINE 
EARL and maybe other relevant languages, and eventually to come up with 
a more thought through specification.

A core feature of the approach in developing the HUMAINE EARL has been 
to let use cases and the resulting requirements drive the specification. 
I suggest to follow the same general approach here, because:

* writing up your own use case(s) makes you think carefully about what 
you expect from an emotion representation language;

* having to justify features in the spec with requirements from use 
cases naturally helps to keep the complexity of the spec in certain limits.


I propose the following approach to the work of the group over the next 
year:

1. Marc circulates a draft Use Case Overview as the basis for 
discussion, initially containing a summary of use cases and requirements 
as proposed in HUMAINE. Starting from what is already there helps us to 
avoid "re-inventing the wheel".

2. Everybody writes concrete use case(s) describing their own work, as a 
contribution to an enriched Use Case Overview document.

3. Discussion on the requirements resulting from the enriched Use Case 
Overview document. If possible, agree on a set of requirements to be 
addressed by an emotion markup language, and on the limits of the types 
of information that should be or should not be contained in that 
language. If such agreement is not possible for all aspects, document 
the disagreement and reasons for disagreement, and proceed.

4. Critical investigation of the HUMAINE EARL spec (and potentially 
other relevant languages) in the light of the requirements documented in 
step 3.

[5a. Formulation of a revised specification.]
   or
[5b. Discussion of various options and reasons for preferring/not 
preferring them.]

How far along this line we will be able to proceed depends entirely on 
the group dynamics, i.e. the amount of time and enthusiasm devoted to 
the issue by all participants. Whether and how we can get to step 5 also 
depends on the level of agreement in the group. Remember that this 
really is an open-ended process: as an Incubator group, we merely 
investigate the feasibility of standardisation, which may or may not be 
given. It is more important to have all relevant view out in the open 
than to agree! :-)

Technically, we have committed to writing a final report [5] by 10 July 
2007. The above steps can actually be conceived as different stages of 
writing that report.

The public mailing list, as well as the phone conference on Monday 
(agenda to follow), provide the first opportunities to voice your 
opinions on this proposed rationale.

Best wishes,
Marc
Initial Chair of the Emotion XG


[1] http://emotion-research.net
[2] http://emotion-research.net/earl, specifically 
http://emotion-research.net/earl/proposal
[3] http://www.w3.org/TR/emma
[4] http://www.w3.org/TR/speech-synthesis
[5] http://www.w3.org/2005/Incubator/XGR/about.html

-- 
Dr. Marc Schröder, Senior Researcher
DFKI GmbH, Stuhlsatzenhausweg 3, D-66123 Saarbrücken, Germany
http://www.dfki.de/~schroed
Here. Now. Real, first-person experience. Am I there to witness it?

Received on Wednesday, 30 August 2006 13:16:38 UTC