Matthew, (01)
Here are responses to your key points. (Full text of Matthew's
comments below). (02)
Jim S (03)
----------------------------------
Paper: "Developing Larger Domains does not Scale. (04)
MW: I dispute this. I have recently finished creating a data model for (most
of) Shell's Downstream business... (05)
JS: The paper agrees larger and larger domain models can be built, which you
have proven, but aserts it gets harder and harder to cover more and more
domains, and eventually becomes infeasible. It goes on to say. "If it were
viable to develop ever larger domain data models, the world could develop one
huge data model and data interoperability would be solved." (06)
--------------------------------------------
MW: In the candidate technical solutions section 2) it is stated that "An
upper data model (for all domains) has never been developed, nor is it feasible
to do so." This is not true. ISO 15926-2 is just such a data model. (07)
JS: The paper defines an ontology and data model and says a data model (for all
domains) can't be built. ISO 15926-2 does not fit the paragraph's definition
of a data model, so the statement is accurate. ISO 15962-2 is closer to an
upper ontology, which the paragraph says has been done. (08)
--------------------------------
MW: ISO 15926-2 is designed to be a cross domain integration data
model/ontology that can be extended into different domains through reference . (09)
JS: This may be possible, but isn't ready yet for implementation by enterprises
from other-than-oil fields. Are there any demonstrations (with published
results) of it being implemented by a totally different domain (e.g. medicine,
entertainment, etc.)? The paper could list it as an example of a single upper
ontology, along with SUMO, Upper Cyc, DOLCE, and BFO. (010)
------------------------------------ (011)
>
>
>
> -----Original Message-----
> From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
> [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of
> matthew.west@xxxxxxxxx
> Sent: Saturday, March 17, 2007 8:02 AM
> To: ontolog-forum@xxxxxxxxxxxxxxxx
> Subject: Re: [ontolog-forum] New paper from SICoP CDSI WG
> ''InteroperabilityAcross the Enterp rise -- Why Current Technology
> can't Achieve it"
>
> Dear Jim,
>
> > Please see attached paper from the SICoP Cross Domain
> Semantic
> > Interoperability WG.
> >
> > The Abstract reads:
> >
> > "Enterprises need data interoperability across all of
> their own
> > systems, and with external systems. The Federal CIO Council
> Strategic
> > Plan, FY 2007 2009, calls for, interoperability across Federal,
>
> > state, tribal, and local governments, as well as partners in the
> > commercial and academic sectors. [1]
> > Current technologies such as XML, metadata, RDF, OWL, and
> stand-alone
> > ontologies can achieve data interoperability, but only within
> domains
> > or Communities of Interest (COI), or between limited numbers of
> these.
> > These technologies cannot today achieve data interoperability
> across
> > the many domains
> > found in most large enterprises. To achieve this goal,
> > leading organizations will need to invest in emerging technologies
> and
> > mature them to where they are ready for enterprise-wide
> > implementation."
> >
> > The CDSI WG web site is at
> > http://www.visualknowledge.com/wiki/CDSI.
>
> This is the space which is the focus for the work I do and which ISO
> 15926 is designed to support, so let me make a few observations on the
> interesting paper you have circulated and a brief overview of where we
> are with ISO 15926 in being able to support this kind of requirement.
>
> Developing Larger Domains' does not Scale
>
> "For example, a Finance and a Logistics community could coordinate to
> develop a large data model, but this would not solve interoperability
> with systems from other domains, such as Acquisition or Human
> Resources."
>
> I dispute this. I have recently finished creating a data model for
> (most of) Shell's Downstream business (taking oil from a tanker,
> turning it into product and delivering it to gas stations). This
> included 5 or 6 areas of similar size to the ones mentioned above.
> We found much commonality across the different domains which only
> needed to be done once, and thus as additional domains are added the
> incremental cost of addition falls. Indeed we know that if we added
> Shell's Upstream business (getting oil out of the ground and putting
> it into oil tankers) there would be very large scale reuse.
>
> Just to give a benchmark the data model has some 1700 entity types and
> cost US$1m to develop over 12 months. It uses ISO 15926 as its
> foundation (of course).
>
> This is not to say that developing such a data model is easy. Unless
> you take a strong ontological approach, then the kinds of problems you
> point to are significant.
>
> Another point worth making is that this is a Conceptual Data Model,
> not a design for a single database that all systems would use. This
> might be theoretically possible, but is not currently practical and is
> probably also not desirable.
>
> In the candidate technical solutions section 2) it is stated that "An
> upper data model (for all domains) has never been developed, nor is it
> feasible to do so"
>
> This is not true. ISO 15926-2 is just such a data model.
> It should be noted though that unlike most data models, entity types
> represent what the paper descibes as "discreet concepts" rather than
> "labels for complex concepts", i.e.
> the entity types have a low attribute count, with information being
> carried in relationships to other entity types. I would agree that
> complex entity types are a barrier to reuse and interoperability.
>
>
> So how does the work of ISO TC184/SC4 support the objecgtives in the
> paper?
>
> The first piece is in ISO 18876 (parts 1 & 2) which provides an
> architecture for interoperability, based on mapping between data
> models (which I agree is a critical component). This shows that one
> data model can be an integrating data model for a number of
> application models, and that one integration model can also act as an
> application data model and be integrated with others. This allows for
> a simple hub and spoke, hierarchical arrnagement, or even peer to peer
> integration.
>
> ISO 15926-2 is designed to be a cross domain integration data model/
> ontology that can be extended into different domains through reference
> data - ISO 15926-4. A reference data library for process engineering
> is in the late stages of development, and there are already moves to
> extend its use into the defense area. This provides a hub ontology
> into which other data models/ontologies can be mapped.
>
> The next key aspect is how to integrate applications. This is the
> business of ISO 15926-7, also in the late stages of development.
> This provides a methodology for defining "templates" where a "short
> hand template" would correspond say to database record structure, but
> mapped by a "longhand template" into the ISO 15926-2 data model and
> ISO 15926-4 Reference Data Library to make their meaning unambiguous,
> and to give a mapping via the hub into other templates that use
> different combinations of data.
>
> There is no magic here. You have to build the templates manually.
>
> The last piece of the jigsaw is the Facade. This is a system to which
> systems behind the facade publish their data to according to the
> templates to make them available outside the facade to authorised
> users using standard web based technology. Thus providing not only the
> potential for intra-enterprise integration, but also inter-enterprise
> integration.
>
> So we have a (nearly) standardised way that information from different
> systems can be made available with a single definition across those
> systems to those authorised to access it.
>
> So what is left? I think the use case provided in the paper is
> interesting.
>
> "An officer on the battlefield is directed to halt one course of
> action and pursue another objective, which had not been anticipated
>
> and for which there is no plan. He has 15 minutes to issue new
> orders. He spends 5 minutes plugging new parameters into his computer
> system, which activates scores of agents to access thousands of data
> sources (many of them unanticipated sources found through searches).
> After 5 minutes, his system presents 3 alternative plans, giving him
> 5 minutes to make the decision. To achieve this, automated agents
> searched, analyzed, and compared data on items such as weather,
> supplies, costs, training readiness, local customs, and maps, and then
> integrated the findings into the 3 alternatives."
>
> So I think we have what it takes to make the information available to
> be searched, but we do not have the application that the officer would
> use to initiate that search and decide what to search for.
>
> In Shell we do of course have our own planning systems to
> semi-automate the decision making in our business. To me these are
> specialist applications that operate over a large sea of data. I'm not
> quite sure how you would expect to solve this part of the problem more
> generally.
>
>
> Regards
>
> Matthew West
> Reference Data Architecture and Standards Manager Shell International
> Petroleum Company Limited Shell Centre, London SE1 7NA, United Kingdom
>
> Tel: +44 20 7934 4490 Mobile: +44 7796 336538
> Email: matthew.west@xxxxxxxxx
> http://www.shell.com
> http://www.matthew-west.org.uk/
>
>
> > -----Original Message-----
> > From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
> > [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx]On Behalf Of
> Schoening,
> > James R C-E LCMC CIO/G6
> > Sent: 16 March 2007 17:56
> > To: 'ontolog-forum@xxxxxxxxxxxxxxxx'
> > Subject: [ontolog-forum] New paper from SICoP CDSI WG
> > ''InteroperabilityAcross the Enterp rise -- Why Current Technology
> > can't Achieve it"
> >
> >
> > Ontolog,
> >
> >
> > James R. Schoening
> > U.S Army C-E LCMC CIO/G6 Office
> > Voice: DSN 992-5812 or (732) 532-5812
> > Fax: DSN 992-7551 or (732) 532-7551
> > Email: James.Schoening@xxxxxxxxxxx
> >
> >
>
>
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Subscribe/Config:
> http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
> Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
> Shared Files: http://ontolog.cim3.net/file/ Community Wiki:
> http://ontolog.cim3.net/wiki/ To Post:
> mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
>
> (012)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (013)
|