- From: Ken Laskey <klaskey@mitre.org>
- Date: Thu, 5 Nov 2009 07:26:04 -0500
- To: Paulo CG Costa <pcosta@gmu.edu>, <patrick.paulson@pnl.gov>
- CC: <public-xg-urw3@w3.org>
- Message-ID: <02439CAE-A22D-464C-9C99-B816FBB0090E@mitre.org>
Paulo and Patrick,
I believe the URW3 public mailing list has now been appropriately
updated and is ready for use. During the discussion period, the notes
captured by Mike Pool state
Paulo Costa and Patrick Paulson agreed to take responsibility for
driving the discussion forward. At the end of 30 days we want to have
a clear articulation of the goals moving forward and the next steps.
Now with a working mailing list and a chance for each of us to settle
back in after the conference, I believe it is time to start the work
on our goals. The full notes from the workshop discussion are
included below.
I look forward to the continuing discussion.
Ken
<URSW5-notes>
The URSW workshop series - What's next for the Uncertainty
Representation and Reasoning for the SW community?
-Ken went through his notes from an Oct. 2008 presentation
History:
-first workshop in 2005
-URSW 2006 generated a use case challenge:
-identity WWW situations requiring uncertainty and
identify methodologies for addressing them
URSW3-XG working group began March '07
Ken discussed/summarized:
-products:
-an uncertainty ontology to
characterize use cases
-14 use cases illustrating
conditions in which uncertainty reasoning would be useful
-recommendations
-open issues:
-Current status
-Possible venues to continue work:
-See Claudia's Novemer 7, 2008
email
The Future? (see my slides)
-Uncertainty Reasoning formalisms as w3c standards
-represent the output of uncertainty reasoning, e.g.,
represent the uncertainty value but not necessarily standardize how to
do reasoning for specific content.
Discussion
How to proceed?
Bart G.: Should we get a structure for representing mappings, e.g.,
embedded in FOL? - Bart w. Ryerson University, Toronto
-Kathy: That falls under the question of whether
existing standards are sufficient?
-Ken L.: the more we're out of sync with things that were
done in the past, the better case we can make
-Kathy: How
Patrick Paulson, pacific nw lab
-I did work with Open providence people and their approach was to
leave it general enough to be used by everybody, try to avoid XML
Schema vs. RDF wars, keep things at an entity-representation point of
view?
-Kathy: that's suggestive of the Common Logic principle,
requirements/content is at an abstract level and can be implemented in
different language
-Patrick: For the Semantic web we have specific problems.
trevor martin: How far did Umberto get with tapping the EU?
-Kathy: we don't know/
Kathy:
a) how valuable is URSW, do we want a URSW 6? What is the purpose
b) Do we want to go beyond URSW and think about an XG for
standardization?
Ken:
-Note that we needn't follow a W3C XG format to get work done between
meetings.
Kathy:
Yes, XG's are for experimental work, and we could use different
format. An advantage is that it gives you some infrastructure, and
some legitimacy as a W3C and a report comes out of it.
Ken:
W3C outputs has some visibility and we had some cross working
Dave Reynolds:
An achievable near terms goal, represent the results of uncertainty
reasoning. Perhaps we could create a W3C note.
Kathy: Another near term goal: annotating data for use in Naive Bayes
novel, i.e., the inputs and outputs of reasoners.
Ken: If we want a standard, there's nothing like showing a real
problem and why your standard works very well.
Bart: Ontology matching workshop has a contest. Do we need a
competition/challenge
Kathy: yes, a challenge might be a good idea for, for example, ISWC
2010. Perhaps we could use linked data sets.
Mike: Perhaps we could get the old EELD data
Patrick:Challenge could be to combine uncertainty.
Kathy: defining challenge problem would be a hard problem in its own
right
Kathy: How many think it would be a good idea to have another XG,
continuity ? 7 said yes no one said no to the question.
challenge problem: 9, we shouldn't: 2
AC, Microsoft: There's been a lot of work in the uncertainty
community. How much cross work has there been?
Kathy: Emphasis here, as opposed to UAI, is on semantic web,
annotations, new DL, et.
AC: might a challenge problem bring in more people for the UR community?
Kathy: SW group combines rep. and reasoning. UAI became an org. on
its own right. In some respects that has happened to us, some of us
are now on main prog. committee.
Ken: Why did Thomas and Claudia say no to challenge problem?
Claudia: We haven't made enough progress. We need to have more
concrete results and then make a challenge.
Ken: Might we put together a charter for this group moving forward?
and then come up with a challenge in 6 months?
Patrick: The challenge problem needn't be accompanied with a paper,
just need to generate results. But there's a chicken and egg problem,
we need a representation to do it and that will cause questions, etc.
Kathy: One possibility would be to drive towards a W3C note. We
could put that together for an XG. Lay out what needs to be standard.
Ken:We probably can't use 'W3C note' now -- deprecated term -- but, it
would have to be an XG report which has equivalent standing.
Kathy: We'd like there to be a goal and some vehicle for formalizing it
Dave R.: Yes, it may have to be a user submission.
Ken:the important point is that we need a crisp description of what
our goals are:
Kathy: We seem to be moving towards a consolidation of what we want to
communicate: (1) We pick an idea, and annotate it. We specify what
needs to be represented and conveyed. (2) We specify an evaluation
and annotate it.
Ken:Perhaps we could simply put together a mailing list, with a goal
of putting together the goals of this group moving forward.
Bart:One suggestion might be to go to the Ontolog forum.
Kathy:Let's not open that up on the Ontolog forum. But let's nail
down what we're doing and invite people to participate.
Bart:yes, let's just create a track on Ontolog.
Kathy:Yes, that's less stuck on W3C, we could also use a GMU mailing
list or Google Groups.
Bart:Open Ontology Repository is a way for people to get together and
specify what they need. Perhaps that's where we could talk about what
is needed.
Kathy:There's an ISWC OOR meeting here tomorrow.
Mike: The EELD challenge problem might be a good place to start for a
challenge problem. (There was then some discussion of whether or not
the data had been annotated with uncertainties.)
Ken:Continue the discuss in 30 days in an email group. Do we have one
or more volunteers to drive the conversation
Paulo Costa and Patrick Paulson agreed to take responsibility for
driving the discussion forward. At the end of 30 days we want to have
a clear articulation of the goals moving forward and the next steps.
We should get the participant list nailed down within the next week,
i.e., circulate the existence of the group to a broad group of
potential participants.
</URSW5-notes>
-----------------------------------------------------------------------------
Ken Laskey
MITRE Corporation, M/S H305 phone: 703-983-7934
7515 Colshire Drive fax: 703-983-1379
McLean VA 22102-7508
Received on Thursday, 5 November 2009 12:26:13 UTC