EMOXG

<Paolo> Scribe: Paolo Baggia

<Paolo> ScribeNick: Paolo

1. Introduction and Purpose

Marc: introduction
... You should know the charter [1] and the result of last year
... You should also know the final requirement document [2]

[1] http://www.w3.org/2005/Incubator/emotion/charter-20071129#deliverables
[2] http://www.w3.org/2005/Incubator/emotion/XGR-emotion

Marc: Paolo suggested to put priority and focus on simple reqs. ... How do we do that?
... Second point should be to write a first draft within three months, then a second one at six months
... In the meantime we might start to discuss it with VBWG (Voice Browser) [3]
... and MMIWG (Multimodal Interaction) [4]

[3] http://www.w3.org/Voice/
[4] http://www.w3.org/2002/mmi/

2. Simplifying and prioritising requirements

Marc: How do we simplify the requirements?
... Also Laurence Devillers is interested in it.

<Bjoern> volunteers for "who"

<Bjoern> Why don't we make use case task forces?

Cathrine: simplify with respect of what?

Marc: relevance

Bjoern: or use cases

Marc: If we start from the use cases, I fear we will re-open the discussions of last year

Ian: Maybe we can go on the requirement and set a value of 1, 2 or 3

<Bjoern> likes 1,2,3 + no opinion

<Ian> must have (1), could have (2), optional (3)

Hannes: I introduced difficult levels during the compilation of use cases

<Ian> I found the wiki very difficult to figure out.. and amend

Hannes: In theory there was the idea of the people to comment on it.
... I could add a new table in the Wiki for the comments.

Marc: Starting from the Wiki document is difficult, I'd prefer to go one by one and assign values to single requirements

Marc: Hannes, would you circulate a summary from the Wiki document?

Hannes: ok

<scribe> ACTION: Hannes to circulate short summary from Wiki doc [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action02]

Bjoern: add value for "no opinion"

<Ian> must have (1), could have (2), no opinion (3)

Paolo: Other W3C WG used another scale for priority of the requirements

Marc: Paolo, would you circulate it?

Paolo: yes

<scribe> ACTION: Paolo to circulate priority scale for the requirements [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action01]

<Ian> w3 previous spec ranking example -> http://www.w3.org/TR/voice-dialog-reqs/

<Ian> must have, should have, nice to have

<Ian> ok

<Bjoern> ok

<Marc> Suggest to first set up a questionnaire,

<Marc> ... getting people's opinions of what is high or low priorities

<Marc> ... and then in the next phone meeting, decide how we proceed

3. Options for an EmotionML (XML, RDF, OWL)

<Bjoern> likes XML format

Marc: This is the most interesting part, but difficult
... we don't know all the options.
... I'll go for XML, who is an expert of other options, e.g RDF?
... Then other decision is between flat or deep structure

Marc: opinions?

<Bjoern> likes deep xml

<Ian> OWL -> http://en.wikipedia.org/wiki/Web_Ontology_Language

<Ian> RDF -> http://en.wikipedia.org/wiki/Resource_Description_Framework

Paolo: Experts should highlight strenght or weaknesses of the different options.

Ian: It depends on the application

Marc: We should consider different applications

Ian: All of them can be used, but it is important to understand what we want to do with the language.

Bill: I worked with OWL, I love your reqs document, I don't know very much of XML
... There are many pre-existent onthologies to express time constraints
... There is a lot of predefined knowledge in OWL

Marc: I know OWL/RDF very little. I cannot express an opinion.

<Ian> we need to "bone up" quickly

Bill: I could dig around on pre-existing examples or write a new one.
... OWL stands on top of RDF and XML.

<Ian> define the parameters

Ian: about parameters, I think the languages should be easy to read
... So one parameter might be readability. We should set general parameters
... then look to the options and select the more fitting one.

Marc: We could first look at the properties of these languages

Hannes: I'd like to understand what we are talking about.
... We should distinguish syntactic and semantic aspects.

Marc: Human will write by hand, e.g XML, while OWL is for automatic reasoning

Ian: Another relevant parameter for protocols is if the language is lightweight or heavy.

Paolo: Another parameter is to be easy to integrate within another language, i.e.
... SSML for TTS and VoiceXML for ASR

Ian: for integration XML is better

<Marc> We might start from a concrete even if very simple example, like EARL

<Ian> Start with simple XML and then expand out as necessary

<Ian> EARL is great actually

<Marc> http://emotion-research.net/earl

<Ian> ok, we can go home now... :)

<Bjoern> likes starting with EARL on an xml basis and only expand if necessary...

<Marc> Design principle for EARL was: "Simple cases should look simple"

Enrico: EARL could be a good starting point.

<Bjoern> scribe: Bjoern Schuller

<Bjoern> scribenick: bjoern

<Ian> i think in the next week or so we should all become basically familiar with the pros and cons of RDF/XML/OWL

<Ian> I like to start with EARL

<Ian> its easier to start from something existing then adapt

Marc: not needed to push it through, but just start based on EARL/XML

<Paolo> Scribenick: Paolo

<Ian> nice

Bill: I can do an example in OWL, while other people can do XML or EARL.
... Then we can compare

<scribe> ACTION: Bill to select part of requirements and to do the example in OWL [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action03]

<Ian> can anyone else do RDF?

<Marc> Do we have anyone with experience in RDF?

<Bjoern> no...

<Ian> I can try

<Ian> yes... !

<scribe> ACTION: Ian to do the example in RDF [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action04]

<Ian> right

<Ian> we should consider the same part

<Myriam> I can try also

<Ian> Bill, I will ask your advice too

<Myriam> with another portion

<Myriam> ok

<Myriam> rdf or owl

<Ian> what about the same portion in XML?

<Bjoern> could do in xml

<Myriam> yes

<Myriam> we start with EARL

Marc: the easiest is that you (Bill, Ian, Myriam) to communicate offline and select the same part.

<Ian> ok

<Myriam> with flat or deep structure?

Marc: is it flat or deep?

Bjoern: probably flat

<Bjoern> does flat xml structure example.

<scribe> ACTION: Bjoern to do the example in XML with flat structure [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action05]

<Myriam> what do you want me to do?

<Ian> deep

<Myriam> ok

<scribe> ACTION: Myriam to do the example in XML with deep structure [recorded in http://www.w3.org/2007/12/19-emotion-minutes.html#action06]

<Ian> do we understand what "deep" and "flat" means?

<Bjoern> yes.

<Ian> ok

<Myriam> flat structure impliees the use of XML attributes

<Ian> ok, thats what i thought, thanks

<Myriam> and deep structure implies simple element or complex element...

4. Date of next meeting

<Bjoern> until 10.01.2008

Marc: deadline should be longer, not realistic with holidays in between.
... next call will be Wednesday very early.

<Ian> any day is ok for me

Marc: week Jan 28?

<Ian> ok

<Myriam> ok

<Bjoern> ok

<Myriam> yes

Enrico: ok

Marc: Work done for the meeting.
... I'll do the poll for the day

Closing part

Marc: Thanks to all the participants!

<Ian> Thank you, good to get back to it

<Myriam> Happy Christmas

Marc: Best wishes and happy new year.