Patrick (01)
ISO 10303 has taken a piecemeal approach to the well understood problem
of communicating engineering data between professional organizations. It
also has established mechanisms to verify that interoperation can happen
accurately, though these only apply to a subset of the concepts it uses.
In one of the major areas of use - product data management - there are
no common concepts, although there are common terms. The meaning of
those terms has been to be negotiated on a case-by-case basis, depending
on the industrial processes involved. (02)
So far it has taken about 30 years to develop, and the development
investment for the standard alone was estimated at $500 million several
years back. It is in production use on a number of major engineering
platforms, and has been mandated by the DoD for use in relevant areas.
Most users are unaware of it because it works - day in, day out. It is
also become the basis for long term data sustainment - engineering data
for platforms (aircraft, ships etc) is expected to have a life cycle of
forty to one hundred years. (03)
I'm sure Matthew could come up with some equally significant figures for
ISO 15926. (04)
If you want traction, you will need to persuade players at level of
government agencies and major industrial players that what you are
proposing is well founded and doable, and will build on existing
capabilities. If you want a brick wall, then try telling them they need
to start again. (05)
See also inline (06)
Sean Barker
Bristol, UK
> -----Original Message-----
> From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
> [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Patrick
> Cassidy
> Sent: 17 February 2010 19:49
> To: '[ontolog-forum] '
> Subject: Re: [ontolog-forum] Foundation ontology, CYC, and Mapping
>
>
> *** WARNING ***
>
> This message has originated outside your organisation,
> either from an external partner or the Global Internet.
> Keep this in mind if you answer this message.
>
>
> Sean,
> The FO project as I have suggested it would include the task of
> finding the set of primitives that would allow logical descriptions of
> the elements in the domain ontologies that are being linked to the FO
> during the testing phase. The use of the (multiple) meanings for the
> words of the Longman defining vocabulary is solely a "preemptive" tactic
> to try to include all the other primitives that might be needed as
> *additional* ontologies are linked to the FO. But even that is unlikely
> to create a complete list - the goal is just to try to anticipate what
> new primitives might be needed so as to reduce the likelihood of the FO
> needing extensive supplementation as time goes on.
>
> But:
> [SB] > If you want to define a set of "primitives", one would need to
>>
>> a) demonstrate that complex ideas can be generated from the primatives
>
>> by the addition of predicates which would allow support the semiotic
>> inferences needed
>> b) demonstrate that the whole system of predicates would not lead to
>> an inconsistent or incomplete understanding of the world;
>>
>
> [PC] I have no ambition to include a *complete* understanding of the
> world, and being as that seems impossible I am not sure what you mean in
> this context. As for demonstrating that complex ideas can be generated
> from the primitives, that is simple for test cases; but I don't know
> what you mean by supporting semiotic inferences. Perhaps an
> illustration? (07)
[SB2} The process of making a semiotic inference would invoke general
background knowledge, possible through tasks such as probabilistic
reasoning. Compare "The car stopped at the STOP light" and "The car
stopped at the RED light". In the first, the semantics associated with
the sign are explicit (STOP-light=ON => STOP-car). In the second, one
would have to identify the various types of red light (fault indication,
immoral-activities-available, dark-room-in-operation, etc) and decide
whether had any associated behaviours that relate to the car stopping. (08)
> [SB2] In this context, by *complete* I mean not hiding or blocking any
> knowledge encoded in for any particular primitive. Apologies, I can't
> think of a good example for the moment.
>
> [SB] > >
>> I suspect that what you want is a decent data exchange standard, of
>> which there are already hundreds, if not thousands. More specifically,
>
>> where you need to start is the big, integrative standards, such as ISO
>
>> 10303, ISO 15926, GJXML, ....
>>
>> If you could start by unifying the first three, that would be a great
>> help :-).
>
> [PC] Those would be standards that would be very important to include
> in the FO project, as well as dozens of others (though I think the NIEM
> might suffice rather than GJXDM). But I do not expect that any small
> project linking two or three such standards will have much effect, and
> even that could not be done convincingly by one person... (09)
[SB2] The fact that the very big standards have difficulty making
headway, even when people can write down the benefits, is an indication
of the uphill struggle needed (010)
> ...In addition to demonstrating that a wide variety of different local
> standards can be integrated using the FO, I think it will be necessary
> for the users to create and make public different applications based on
> ontologies or databases linked to the FO, and demonstrate that those
> applications can accurately interoperate... (011)
[SB2] This is something that terrifies the application providers, since
it means that you can use somebody application. Hence they avoid it like
the plague, except where various of their major customers gang up in
angry meetings (and I mean, very major corporations). (012)
> ... At this point I think that the most vital thing missing is **public
> open-source** applications using ontologies to drag the topic out of the
> etheral realm of theory into the real world for the vast armies of
> potential users with no interest in the details of ontology
> construction. I also think that without a NL interface t
> he result will be very difficult to use. This requires considerable
> work. On my own I do intend to do some small tests of the FO principle
> (progressing slowly), but do not expect anyone to be convinced by that.
>
> I think that a centrally controlled multi-group effort is necessary -
> or perhaps something will be forced on us by Google or Microsoft. I
> also expect that an FO will eventually be adopted, as the rationale is
> very compelling to me. But *how* we will get from here to there could
> go by a number of different paths. I just think that the FO project is
> the most direct, and fastest. I would be intrigued by any other method
> suggested to get to the same result quickly.
>
> Pat
>
> Patrick Cassidy
> MICRA, Inc.
> 908-561-3416
> cell: 908-565-4053
> cassidy@xxxxxxxxx
>
>
> (013)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: 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 join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (014)
|