ppy/chat-transcript_20090319d_edited.txt
PeterYim: Welcome to the OntologySummit2009 Panel Session - "Standards Community Perspectives Toward Ontology-based Standards" - Thu 19-Mar-2009
Subject: The Standards Perspective -- What are standards and how could ontology help?
* Session Chair: Mr. HowardMason (ISO TC 184/SC 4; BAE Systems) -- "What do standards need?"
* Panelists:
o Dr. MatthewWest -- "Ontologising Standards the low hanging fruit"
o Mr. DavidLeal -- "Ontologies derived from standard information models used for engineering analysis meta-data"
o Mr. GeraldRadack -- "Catalogues, Dictionaries, Libraries, Data and Metadata"
o Mr. NicolasFigay -- "Semantic PLM: using OWL for semantic repositories and mapping using reasoning engines"
o Mr. DavidShorter - "Seeking greater rigour in textual standards"
o Mr. DavidPrice - "Future STEP"
Refer to details on the session page
at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2009_03_19
anonymous morphed into NicolasFigay
anonymous morphed into DavidLeal
PeterYim: Hi you are at the right place ... session starting in 5 minutes
anonymous morphed into MikeAxelrod
anonymous morphed into DavidPrice
anonymous morphed into Matthew West
anonymous morphed into DavidShorter
NicolasFigay: Some reaction:
NicolasFigay: EXPRESS OWL P21 are not only formats, but also rely on different modeling concepts
and paradigms. Not so easy to make it coherent
MikeBennett: The first step I would think is to have a common concept of layers of types of model
(as found in computer model driven development) so that logical models
and semantic models are seen as distinct.
DavidPrice: In reply to Nicolas comment, I think it's actually UML, EXPRESS and OWL that need alignment.
Agree that paradigms are different, but not so different that a mapping (perhaps lossy)
is impossible, or at least with a bit of additional information/annotation of the models,
schemas and/or ontologies to support the mapping.
DavidLeal: URIs for country codes seem an obvious place to start for the very simple stuff - much easier
than units of measure. ISO 3166-1 says Afghanistan is "AF", and the codes are already on the Web.
Assigning a URI such as http://www.iso.org/iso/country#af seems obvious. To complete the task,
the ISO 3166 Maintenance Agency needs to assign a URI to the class Country too.
DavidPrice: In reply to DavidLeal comment, while URIs for country identification is good - engineering apps
require units far more often so while that may be harder, it has far more value.
anonymous morphed into DeborahMacPherson
DavidPrice: By the way, in order to enable better alignment between STEP information models and ontologies,
while we are not changing EXPRESS itself, we are gradually adding capability to the information models
themselves (e.g. "Classification Assignment By External Class" and a new "Same As External Item" I've
just created as part of AP239 E2 work).
JoelBender: a significant amount of work in units has already been done with/in NIST, it needs to be unified
with the already published concepts and terms of the BIPM.
DavidLeal: I agree with DavidPrice, but it would set a precedent. I also agree with Joel.
DavidPrice: It would be great if we could push that NIST/BIPM units activities ... any idea how?
DavidPrice: The concerns Nicolas described align very well with the issues/concerns/approach that
we're working with NIST on the Future STEP project.
SteveRay: Perhaps having this forum conclude that an ontologically-driven alignment between the NIST units work
and BIPM should be on the short list produced at the Summit on April 6-7 would be a good stake in the ground.
JoelBender: w00t!
NicolasFigay: What is BIPM?
JoelBender: http://en.wikipedia.org/wiki/International_Bureau_of_Weights_and_Measures
DavidLeal: I have raised this with ISO TC12, responsible for the ISO 80000 series which is replacing ISO 31 and
ISO 1000. They need to be involved too.
DavidPrice: Steve, that would be great.
MikeBennett: Do people using SC4, SC5 etc. always mean TC184? I work on TC68/SC4 and there must be others
once we reach out to other industries? SC is Sub Committee.
SteveRay: For those present today, generally TC184 is understood. (Implicit knowledge!)
JoelBender: If any of the folks working on this are familiar with MathML, that would be a plus. It's a fairly
straight forward process to translate MathML terms into rdf/rdfs/owl constructs,
and that's the one spot where the NIST work stops and need to keep going.
NicolasFigay: I would recommand not to make evolve EXPRESS too fast, and not to be align with other language
which have not the same focus. If not, it will become a monster, and we will have difficulties
to choose one language or the other for a given purpose. If EXPRESS expressivity is the same than OWL,
it is a kind of competition. If different, it is a cooperation. The exemple is the way RDL are used in PLCS,
that is a little bit difficult for non expert
DavidPrice: There is no plan to change EXPRESS.
NicolasFigay: Ouf
MikeBennett: @ Steve, thanks, I see that's Automation systems and integration so I presume that there are opportunities
for standardisation using ontologies in other inustries e.g. TC68 financial services where we are
doing things in this area. So I did not have the explicit knowledge that you mention
DavidPrice: ISO 19440 that DavidShorter is describing looks very much like MODAF/DODAF.
SteveRay: Mike, There are indeed opportunities. I will follow up on your email to me suggesting the financial industry.
Also, what are your thoughts on XBRL?
DavidPrice: I should have said, a very small subset of MODAF/DODAF.
NicolasFigay: To respond to Mike, we are also involved in System Engineering, Enterprise modeling, project management,
Business Intelligence and trying to make links ... through ontologies and components agregation
MikeBennett: @ Steve - XBRL very important. I am waiting on the OWL version so I can use it more formally
in creating the financial securities ontology. I've referred to it as much as I can, and I think
everyone involved in SC4 (gotcha!) and elsewhere acknowledges its importance in this area.
MikeBennett: @ Nicolas, thanks, I wasn't aware of that. Are you also aware of UN/CEFACT CCTS and so on?
NicolasFigay: To replace Protg, you can use a UML modeler, with ODM profile but compliant with XMI!
MikeBennett: David, great point about UML and publishing something viewable. I've been using an ODM based format
(I see Nicolas beat me to it)
MikeBennett: But with different coloured archetypes etc.
NicolasFigay: Don't forget configuration managment of language format
DavidPrice: My cell phone won't push the #3 to you, so I cannot talk at the moment.
DavidPrice: I am doing #3, but you still cannot hear me.
JoelBender: *3
SteveRay: Mike, is XBRL - OWL version being worked on already?
MikeBennett: So I heard, but I haven't heard anything about this recently.
NicolasFigay: I don't like the idea of migration. May be due to Long Term archiving concern...
NicolasFigay: For ABOX and TBOX, it is not only express but also p21
NicolasFigay: For these mapping, the configuration management is to be carefully handled for different languages,
conformance and openess of used tools (e.g. use only tools providing correct and interchangeable XMI
with appropriate version). Horizon and rythm of evolution should be compatible and align
- some synchronisation of the communities is required
NicolasFigay: A tool can be the set of framework being produced by IAI, ASD SSG, ISO ... for governance of standards
... which should include those related to ontology
NicolasFigay: still, it is a pity not to have semantic services for SOA.
NicolasFigay: Business Process are not to be forgotten
JoelBender: Will SKOS be a unifying force?
DavidPrice: I have to leave to present the Future STEP project o PDES inc. Sorry!
MichaelGruninger: What are some of the terms that would be included in an ontology of product structure?
NicolasFigay: From my point, it will be extension of modeling constructs that should be added
JoelBender: I found this which is a great resource
for understanding why discussions of "is a part of" are hard to manage
(and why OWL classification fails sometimes) particularly for me with non-english speakers.
MikeAxelrod: I've managed to unmute'
SteveRay: I have begun to list the ideas of candidate standards on the conference page for next week's panel,
under "Agenda Ideas", see http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2009_03_26#nid1UG4
MichaelGruninger: There are already several first-order ontologies for mereology (axioms for part-of) and mereotopology
(axioms for part-of and connection). What are other relations that need to be axiomatized for
product structure within the current standards?
NicolasFigay: if automatic transformation language to language is not very usefull (impedance mismatch),
we should find a way to capture the way it was translated in order to be able to go back
- current technology manage it (with annotations
JoelBender: @ MichaelGruninger: where can I find these?
NicolasFigay: owl, uml, java ...
NicolasFigay: you are welcome
MichaelGruninger: I will provide a complete bibliography on the wiki. We also plan to have an initial version of
JoelBender: excellent...thank you...
NicolasFigay: Thanks Howard
SteveRay: Great session! Thanks to all.
PeterYim: great session ... the audio recording and chat transcript will be up on the session page later (as usual)
MikeBennett: Some great ideas in there. I thought there was a lot of valuable detail in what Nicolas is doing for example.
PeterYim: bye everyone ... thanks you for the participation!
PeterYim: posting comments received (on email) from Jerry Smith (DISA) below ... {nid 1VI4}
GeraldSmith: Amplification Gerry Radack comments on opportunities for quickly transforming and elevating a specification:
SC4 "Harvesting" is a fast, low administrative drag process for progressing a candidate specification
to international recognition and accreditation.
Source of the specification is immaterial as long as quality, IPR and management issues are resolved.
See http://www.tc184-sc4.org - go to the "Harvesting" link. {nid 1VI5}
GeraldSmith: Regarding DavidShorter's question on his last slide about "... how the result should be published as a standard?",
I would suggest all to look at the "Harvesting" process of ISO TC184-SC4. ... See: http://www.isotc184-sc4.org {nid 1VI6}
PeterYim: I look forward to all of you to send me (via email to ) your support
by confirming that we can list your standards working group and/or your organization
as a "co-sponsor" for this Summit
- ref.: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2009#nid1Q4P {nid 1VI7}