Video/Media WG proposals - feedback and comments (IG dependency; typos; scope scope & scope :)

Hi there

I'm checking out the new Media/Video proposals, 
http://www.w3.org/QA/2008/04/proposed_activity_for_video_on.html

A few notes on the docs here:

http://www.w3.org/2008/01/media-fragments-wg.html
$Date: 2008/04/20 07:31:19 $
[[
Semantic Web Education and Outreach (SWEO) Interest Group
     This Interest Group has been established to develop strategies and 
materials to increase awareness among the Web community of the need and 
benefit for the Semantic Web, and educate the Web community regarding 
related solutions and technologies. The Cool URIs for the Semantic Web 
is one of the Interest Group Notes produced by the Group.
]]

vs

http://www.w3.org/2001/sw/sweo/
[[
This Interest Group is now closed: its official charter expired on the 
31st of March, 2008. Note, however, that some of the activities started 
by this Interest Group have not been stopped, but go on independently 
instead.
]]

You might consider a reference to the Semantic Web IG, assuming a 
charter extension beyond april 2008 goes through OK (which was last plan 
I heard via Ivan).

http://www.w3.org/2001/sw/interest/
http://www.w3.org/2006/07/swig-charter.html [[ The Semantic Web Interest 
Group is part of the W3C Semantic Web Activity and is chartered until 
the end of the Activity, 30 April 2008. ]]


In http://www.w3.org/2008/01/media-fragments-wg.html
under "1.1 Success criteria",

"Ability to link into video, audio, and images content"

I suggest instead, "Ability to identify and link into video, audio, and 
images content"

...as this emphasises the bridge to metadata/annotation concerns. From a 
metadata perspective, if I know that movie_1.mpg is a concert featuring 
a certain band, and that from 12m:41s to 14m:00s is a particular track, 
these media-fragment identifiers should be enough to allow quite rich 
external metadata about the performance. Using only the word 'link' 
emphasises hypertext scenarios over metadata ones. I realise you have a 
whole other WG proposed for metadata, ... but in the interests of the 
fragments group understanding that their remit also covers metadata 
implicitly, I suggest this change.

http://www.w3.org/2008/01/media-annotations-wg.html doesn't seem nearly 
as well scoped or drafted; which is understandable as it's a massively 
more complex problem space.

Under 1.1 Success Criteria in the annotations WG draft, you have:

"Abilitity to convert metadata information from one metadata standard to 
an other"

This is a noble if vague goal. I suggest specifying some specific target 
standards.

Minor edits: "abilitity" -> "ability"; "SparQL" -> "SPARQL"
typo: representating

Under "1.2 Out of Scope", the draft just as "Anything to have here?"

I suggest making it clear there whether the WG scope is discrete/finite 
media "objects" or whether (potentially un-ending) streamed video is 
also in scope. My instinct here is that streamed out to be in scope, 
both live-streamed and on-demand.

Where you have "A W3C Recommendation describing a simple ontology for 
representating a set of concepts often associated with media objects."
...you ask a key question:

[[
   How to decide which concept to include in a "simple" ontology? Should 
we say only 20 concepts? 30? Should we based criteria on use cases?
]]

For a yardstick, remember that the dublin core community began their 
work with 15 properties in 1995, and things are still moving along 13 
years later. I wouldn't try to measure simplicity by counting the number 
of terms. As the draft suggests, use cases might be a better measure 
here. Also in ontology development, 30 terms where 25 are properties, 5 
are classes, gives you something rather different to an ontology of 25 
classes, 5 properties. I'd argue that surplus classes are a bigger 
complexity problem than surplus properties, but I wouldn't expect 
everyone to agree here.

Another scope question: the broad umbrella here is "video" yet images 
are mentioned in passing. It would be very helpful if the WG charter 
could explicitly list some media types considered in scope:

eg.: vector images; bitmap images; flash movies; interactive webapps in 
SVG; interactive webapps in W3C widget spec form; or in Google 
OpenSocial form, ... etc. Audio files. Audio streams. Objects in 3d 
worlds. VRML documents. etc.

Any one of these could chew up a good few telecons, and the WG schedule 
draft has them reaching Last Call "we think we're done" stage in 4 
months, July-October 2008. Since W3C WGs are noticably slower in the 
(northern) summer, this looks over confident. Either reduce the scope 
("Version 1 targets only video files on webservers", "Version 1 collects 
only some useful classes for different forms of online media content", 
...) or give the WG another few months (another 6, I'd guess). I can't 
imagine this WG moving as fast as the GRDDL group managed to, to give a 
comparison: http://www.w3.org/2001/sw/grddl-wg/#charter_history ...

Also -in passing- check out the Perl Net::Flickr::Backup RDF output to 
see how many namespaces can be used to backup a 'mere' static photo:

http://search.cpan.org/src/ASCOPE/Net-Flickr-Backup-2.6/README

      <rdf:RDF
       xmlns:dc="http://purl.org/dc/elements/1.1/"
       xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#"
       xmlns:a="http://www.w3.org/2000/10/annotation-ns"
       xmlns:acl="http://www.w3.org/2001/02/acls#"
       xmlns:exif="http://nwalsh.com/rdf/exif#"
       xmlns:skos="http://www.w3.org/2004/02/skos/core#"
       xmlns:cc="http://web.resource.org/cc/"
       xmlns:foaf="http://xmlns.com/foaf/0.1/"
       xmlns:exifi="http://nwalsh.com/rdf/exif-intrinsic#"
       xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
       xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
       xmlns:flickr="x-urn:flickr:"
       xmlns:dcterms="http://purl.org/dc/terms/"
       xmlns:i="http://www.w3.org/2004/02/image-regions#"/> ...




Hope this helps; good luck with the activity! It looks well worth 
starting...

cheers,

Dan

--
http://danbri.org/

Received on Thursday, 24 April 2008 21:51:45 UTC