[Accessforall] Minutes - Context Description meeting - 2012-07-03

Gottfried Zimmermann (List) zimmermann at accesstechnologiesgroup.com
Tue Jul 3 15:03:19 UTC 2012

Dear project partners,


the minutes of today's telecon are posted at
Description and below my signature.


I will inform you about our next meeting, once I have heard back from Dan







 Prof. Dr. Gottfried Zimmermann

Mobile User Interaction



+Description> 2012-07-03 Meeting on Context Description 

Tue Jul 3, 12:00-14:00 UTC

2&hour=12&min=00&sec=0&p1=0> Translate to my time zone


This meeting is being held as part of the Raising the Floor Consortium. The
Raising the Floor membership agreement, in particular its IPR policy,
applies to the contents of the meeting.


*	Christophe, Gottfried
*	Claudia
*	Jim
*	Liddy
*	Vassilis
*	Andres
*	Andy
*	Gregg

Approval of last minutes

+Description> 2012-06-26 Meeting on Context Description approved.

Review of the Context Description table

e=Context+Description&linkCreation=true&fromPageId=31195159> Context

Do we want to give names to context or describe their functional

*	Continuum between giving names and give exact data.
*	E.g. "Home" is not a globally unique environment, it expresses a
person's own home environment. "Home" is a name for a situation rather than
describing the situation in detail.
*	Looking at terms describing the environment of the user using a
*	Useful for disambiguating a situation.
*	Two types of context properties: giving names (idiosyncratic) vs.
data description

How to go about schema.org?

*	Schema.org can provide properties for location & activities
*	AccessForAll Metadata is doing accessibility metadata in schema.org
(learning resources people)
*	Schema.org is search-engine driven
*	We could define a new object "Context", and link existing terms as
fields (e.g. coffee shop)
*	Light-weight alternative: Just use the labels of the classes, e.g.

What to do about conditions that contain context properties that have
undefined values?

*	Environmental data is typically patchy
*	Some property values may be undefined
*	"Try harder" concept: Use a simple match first, and if it doesn't
work, use a more sophisticated matching.
*	Fuzzy logic: compute with undefined values, and if the condition
evaluates to undefined, don't use the preference
*	Alternative: Assume default values. Easier to compute.
*	Device/environment could ask the user for values.
*	Can this be implementation specific? - But it can make a big

*	Should be known by an entity writing a preference into a user

*	We should also connect to the device properties group.

Next steps

*	Have a discussion with Dan Brickley on schema.org
*	People fill in terms in
e=Context+Description&linkCreation=true&fromPageId=31195159> Context
Description table

Action Items

Fill table at  <http://wiki.gpii.net/index.php/Context_Properties>
http://wiki.gpii.net/index.php/Context_Properties with information from
other projects and standards:

*	Vassilis: User Profile Ontology, paper on context modeling, SOUPA
*	Claudia: ISO 9241-11 Annex A
*	Christophe: RFCs 4589 & 4480
*	Andy: Schema.org & ISO/IEC 24751 

*	see  <http://schema.org/docs/full.html>

*	Gottfried: Contact somebody from MyUI
*	Andres: UsiXML & ISO/IEC 24756
*	Gottfried: WHO ICF, ISO/IEC 24752

See also: ISO 8601, XML Schema Definition - Part 2

Future Meetings

Possible dates & times (to be announced by Gottfried ahead of the meeting):

*	Wed Jul 11, 12-14 UTC = 14-16 CET (not available: Gregg, Christophe,
*	Tue Jul 17, 12-14 UTC = 14-16 CET (not available: Gottfried)
*	Tue Jul 24, 12-14 UTC = 14-16 CET
*	Tue Jul 31, 12-14 UTC = 14-16 CET

Future Agenda Items


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.idrc.ocad.ca/pipermail/accessforall/attachments/20120703/3d233f19/attachment-0001.html>

More information about the Accessforall mailing list