[All] Notes from Wrapup Session at F2F

<Mike Uschold Scribing>

Future Plans
PORT TF  Alistair - 
*	Should work continue? Yes. 
	0	Should it be a recommendation? Probably. 
*	There are existing documents
			1	There is a list of issues.
*	Relationship between SKOS-like stuff and owl ontologies, how
co-exist? How migrate from one to other.  Need to know how this works to
do SKOS right. Many different scenarios to look at, other than living
side by side.  Relationship between taggers and formal ontologies.
				1	Community Involvement:  Need to
make sure people coming to this community know where to start. Need
engagement form maintainers and users of SKOS.
				DanBri: this work is needed before other
red-track work starts on this topic.
				2	Multi-linguality - lexical
issues
*	Other work worth doing: YES.
*	Mapping of SKOS-like stuff. (started in SWAD Europe) Maybe take
some existing mapping stuff from other communities, and rdf-ize it.
Multi-linguality also needs mapping.
Welty Wisecrack: Do you mean OWL and RDF are different languages?

WORDNET TF: Aldo
Must finalize draft ASAP. Waiting for reviewers feedback.
Issues: 
*	2nd note on WN uses aimed at application developers 
Guus: goal is to get draft out into community, to get feedback, how can
application developers use it?
*	Community consensus, collaboration with Princeton
*	Make available as Web service
*	Comprehensive lexical markup (link to LNF)

Time Frame to completion: one year after end WG

MULTI-MEDIA TF:
Need extension of WG to finish interoperability document. (need 6 months
extension)
New notes:
*	video (requires new WG or new charter (1-2 years time frame to
do same deliverables for Video)
*	VRA note: try to merge into future SKOS activity ( longer term
issue, depends on community consensus)


TOPIC MAPS: 
Valentine: 
*	Interoperability guidelines document: (6-12 months time frame)

RDF in HTML, RDFA: Jeremy
Not clear how XHTML community can have input from Semantic Web community
after the SWBPD WG disbands. (1 year)

Should be Semantic Web tools that can read XHTML2, notes to write (more
than one year)

SOFTWARE ENGINEERING: Phil

More notes:
*	Compound Keys (6 months)
*	Java implementers (9-12 months)
*	Jeff Pan:  Automated software engineering (1-2 years) experts in
software engineering, how can sw technology be used...
--


XSD: Jeremy

Will finish on time.
Future: input for RDF/OWL revision (one fine day)

ADTF: 
Libby - will finish on time. Input to SWEO

New TF (ish) - TUTORIALS
*	input to SWEO (proposed)


VM - Vocabulary Management [ Alistair ]
Bite sized bits: should finish HTTP URI soon.
*	How  does ... solve expressivity problem. How publish ontology
when different users want different expressivity OWLS lite, dl and full.
Documentation of current practice [ 6 months]
*	Change management / versioning for vocabularies [1 year]
*	Good practice for HTML documentation for vocabularies [6-12
months] Want to tie into OWL documentation. Human readable documentation
for ontologies/vocabularies. 


OEP: 
13 New documents  suggested form people outside OEP ( some may belong to
SWEO).
On OEP Web page.


LEFTOVER IDEAS that may not have been addressed, that were on original
list of things to do for SWBPD WG.

*	use of RDF label, etc
*	tools index (for SWEO?) out of scope for us
*	units and measures
*	ontology design issues (covered in OEP note ideas)

So just one idea to pursue from original list.

NEW IDEAS:
*	study link between conceptual graphs and RDFS/OWL
*	survey existing mappings
			1	find common core of graph family (CGs,
RDFS, RDF etc)
			2	usage note
*	good practice for internationalization features for RDFS/OWL,
what is there, what is missing, how to use it?

LOOKING TO THE FUTURE: how/whether to continue the work being done in
this WG.

Look at different scenarios for extending SKOS work.
*	Separate SKOS WG, and then on to Rec track 
*	Continue in extended SWBPD WG, then on to Rec track

Ralph: why should specific rec-track work like this be in a WG?
Alistair: 1st, to ensure that SKOS work stays in tune and connected with
the ontology community. 2nd: people not directly SKOS folk, but who come
into SKOS also get a link to the wider Semantic Web community.

Two scenarios, to recharter or not to recharter, that is the question.

Re-Charter: max period of ?  Ralph not think there is any limitation on
this. 2 years would be a reasonable proposal.

Different from asking for shortish extension.

For each scenario, how would the work proceed?

Scenario A: do NOT recharter
How to continue SKOS? 
*	rec track
*	link with VRA work

WN has a problem, cannot finish all work in time, may continue in SW
Interest Group (SWIG)
DanBri: this could work, would have to be more visible to wider
community.


SWIG:  xx means not well placed in SWIG
SWIG is a loose open forum, not a tradition of publishing notes in
formal way. SWIG is a community, makes more work for the coordination
group (SWIG chairs). Would be taking over job of Guus and DavidW
*	WN
*	MM Video
*	RDFTM
*	X - RDF-In-XHTML ill fit, because too important to be in just an
IG.
*	SE - rest
Phil: fits in short term, but this area may grow too fast, would fester
in an IG
ChrisW: there is an argument for keeping it here, for sake of outreach
Phil: I agree with that.
*	VM: change management & documentation 
*	OEP  rest

Jeremy: how do we justify to our bosses that we do this work. Harder in
an IG than in a WG. Less of a problem, if there are clear deliverables.
Phil: having TFs in a Best Practices WG, sends a strong signal to the
world, not nearly the same clout, just being in an IG is not nearly as
good.
Guus: I like suggestion that SE notes go into SWEO.

BP Extensions
*	WN, MM interoperability
	0	RDFTM - interoperability 
	1	SE - compound keys
	2	VM: RDF/OWL
	3	OEP
*	Closure axioms
			1	Guidance on domain/range restrictions
*	


Idea: bring OEP under SKOS activity?
Alister: Chris: you belong to Meee!
Discussion that OEP is much broader than SKOS, only a modest sized
overlap.



Scenario B: DO recharter
SWBPD-2
*	SKOS (REC?)
*	OEP- guidelines vs. particular vocabularies/ontologies 
*	VM
*	Semantic Web language core

Guus: need clear technical reason to be in this group.
Phil: link to UML is key
Guss: yes, but link to ODM not needed.
Phil: these are essentially the same thing.
Aldo: relationship between linguistics and Semantic Web is important. It
is under-represented in this WG.

SWIG


Liaison activity seems to be common thread. Is it strong enough for
separate WG?
Guus: Liaison was one of the original goals of the SWBPD 

MM: needs out reach beyond Semantic Web community.
DanBri: scared of having this in SWIG. Other's agree.


New Group Names?
*	Semantic Web Liaison Activity: MM, TM, ODM (SE)
*	Ontology and Vocabulary Management: naturally includes SKOS,
OEP, VM, SWL Core


ChrisW: we need stronger focus, too broad now, hard to get anything done
in telecons
Jeremy: much work is sub-critical mass, becomes possible to continue, if
have the WG, may fall apart otherwise

Alistair: I like idea of focus, as well as idea of liaison.
ChrisW: I disagree with Jeremy, to scattered hinders work
Jeremy: That's true, too.

ChrisW: can we recharter SWEO to add the liaison activity?\
MikeU: great idea!
DanB: that would be too much like SWBPD
CW: no, SWBPD not has not liaison activity.
Guus: I think not appropriate place
CW: we change charter, so it is appropriate
DavidB: I don't think appropriate.
DanB: I don't want to have a SKOS-only WG, too thesaurusy
Jacco: I want technical feedback, SWEO may not be very useful for that
Alistair: what would be products of they liaison work?
	A: Jacco: same deliverables as MM now, but for videos
            A: Guus: same pattern as usual, survey, standards, and usage
guidelines

Stamou: please tell me more about SWL core
A: Guus: place to collect changes, changes, etc of OWL/RDF specs
Stamou: need more datatypes

Guus: MM not fit under ontology and vocabulary management WG.

MikeU: maybe have new WG called Ontology and Vocabulary Management and
Liaison Activity.
Guus: alternate telecons
MikeU: that solves telecon/focus problem

BrianM: classis situation, lots to do, don't try to do them all. Must
have focus. Decide what NOT to do! I propose dropping liaison, it spins
out of control.
Jeremy: clapping...

ChrisW: amounts to where to fit work, organizationally.
Guus: we have gotten much work, spin out of control not actually
happened.

DanB: incubator idea... W3C members can take up forum that does not take
up W3C staff time. This can result in report, which can lead to more
formal outlet.

Q: how differ form SWIG?
A: DanB: good question.

ChrisW: can you use w3C-supported zakim telecon in an incubator?
Stamou: we are putting MM in liaison... there are other problems. Is
part of work of MM that should be done inside Semantic Web area.
Guus: can we cast it as ontology/vocabulary management?
Stamou: advantages of Semantic Web is: you can have more powerful
semantics
Vasilis: there is lots of stuff not just closely related, but the same
thing as Semantic Web. Datatypes, temporal reasoning, ...
Guus: I consider those things that are MM-specific ... if there are
particular representational issues, then they fit in ontology/vocabulary
management.
Stamou: example... MPEG7 - started with new language, dropped some of
their work and took XML Schema, so they might actually pull in RDF/OWL
as well. Needs liaison activity for this.

MM conclusion: two options 1) if we have O/V management group, then the
liaison activity will not be done there. 2) if we have focused BP group,
then can have incubator group for MM

MikeU: this can split the coherence of the MM folk, splinter into two
groups.

DanB: a different approach...more freeform, mailing lists, yahoo groups
whatever, wait to see what comes, see if reports etc are produced. Can
be published through W3C. This is possibly last resort, if something at
risk of falling into the bin.

Guus: in any short extension of this current group, I am willing to
chair it, if re-chartered for a new longish period, it will  be needing
a new chair, I will step down from that role.

Received on Sunday, 6 November 2005 19:22:48 UTC