On 3/6/2012 6:08 PM, henson graves wrote:
> David,
> As you know human artifacts turn out to be useful for purposes which their
> designers never imagined or intended, in this case I am speaking of both UML
> and OWL. (01)
Yes, but they often turn out not to be inappropriate for other purposes too. (02)
>
> The basis for my experiment was that I wanted to see how suitable UML/SysML
> was for chemistry and biomedical applications. I am aware of how OWL and
> extensions have been used in this arena. My take away from the experiment of
> building a water molecule in SysML is that UML is very suitable for these
> applications. The experiment also helped me pinpoint where UML/SysML lacked
> DL expressivity that is really needed in UML/SysML. (03)
I doubt that any chemical engineer would think UML suitable for building
a water molecule - and that's the crux of my comment. They already have
far better tools for doing that kind of thing. (04)
>
> Regarding OWL it sounds like you also have the view that it is for use only
> by ontologists to do whatever they do and by software engineers, but not
> system engineers. Indeed many OWLites may think this way. My earliest
> interest in OWL, following your lead, was to find out if OWL could be used
> for engineering purposes. The answer is that it makes good sense, but there
> are some limitations. The initial result of this inquire is a paper with
> Ian Horrocks, entitled Application of OWL 1.1 to Systems Engineering. Pros
> and cons are discussed. (05)
I do think OWL is plumbing wrt engineering applications. There may be a
few places where it can be used more-or-less directly but they seem to
be few and far between. My 2006 paper did show an interesting use case
but I've struggled to find many more - and I owe you a call on that
topic, a a re-reading of your paper. (06)
>
> Back to the water molecule it is very interesting to compare the results in
> both languages. Both results are partial due to limitations on both
> languages. My reason for doing the experiment was exactly to sort out these
> issues, not advocate for one religion or the other. I hope, based on these
> comments, you can understand why I would do the experiment. Am I free to
> quote that you cannot imagine why I would do such an experiment when I
> publish the paper. (07)
What I think you may have shown is that the meta-model of UML is capable
of representing some, but not all, of the data required to represent a
water molecule, as is OWL - and that's a useful result as far as the
representational capabilities/limitations of these two languages.
Perhaps I've misunderstood, but suggesting that UML/SysML as a
*notation* are useful to chemists does not seem to me to be what you've
shown - and feel free to quote me on that:-) (08)
Cheers,
David (09)
>
> I don't think that your proposed comparison is the correct one.
>
> - Henson
>
>
> -----Original Message-----
> From: ontology-summit-bounces@xxxxxxxxxxxxxxxx
> [mailto:ontology-summit-bounces@xxxxxxxxxxxxxxxx] On Behalf Of David Price
> Sent: Tuesday, March 06, 2012 5:01 AM
> To: ontology-summit@xxxxxxxxxxxxxxxx
> Subject: Re: [ontology-summit] INCOSE Ontology Action Group, onto SysML/UML
>
> On 3/6/2012 4:10 AM, John F. Sowa wrote:
>> On 3/5/2012 3:36 PM, henson graves wrote:
>>> I personally have tasked two
>>> engineers with similar capability to build a model of the molecule
>>> Water in SysML and in OWL. Take a guess about the results. Also guess
>>> how well it can be done in either of these languages or other candidates.
> I don't understand why one would do that. SysML is a tool designed for use
> by systems engineers, not chemists, and OWL is a language for use by
> ontologists and software developers. A more accurate comparison would be to
> have them design something 'engineery' in SysML vs SQL or the SysML
> meta-model or OWL. Few, if any, systems engineering applications would
> present OWL to an engineer as OWL.
>
>> I'm sure that engineers can learn UML or SysML much faster and use it
>> more effectively and more accurately than OWL. But since you did the
>> experiment, I'd like to hear any further details you observed. Did
>> you write a report about them?
>>
>> And I'd also like to hear the opinions of the engineers about which
>> tool(s) they would prefer to use and why.
> 'Engineers' is far too general a term to answer this question.
> Mechanical engineers want to see 3D solid models, EEs want to see circuit
> diagrams, stress or thermal engineers want to see colorful visualizations
> that highlight problem areas, product management/maintenance engineers want
> to see tree-like breakdowns of products and related processes, chemical
> engineers want to see compound structure diagrams, etc.
>
> OWL is just plumbing no discipline engineer would ever see, unless your
> discipline happens to be ontology/software engineering and so you're working
> directly with the pipes.
>
>>> I personally am convinced that HOL in the form of type theory will
> eventually
>>> win out, but this is pretty much irrelevant to putting a stake in the
> ground
>>> with respect to achieving tech transfer. As you have stated start
> wherever
>>> you want, as long as where you start can be given a formal semantics.
>> I agree. But I would avoid using the term "win out" with respect to any
>> particular notation.
>>
>> As I said before, I believe that semantic systems need to support
>> "anything and everything". That would include the Semantic Web tools
>> and many different variations of graphic and linear notations. Then
>> people with different backgrounds can view the common semantics in
>> their preferred notations, and they can use notations that are tailored
>> to their requirements.
> Exactly! Users of an onotology-based application, be it based in OWL or
> HOL, would likely never have that fact made visible to them. We should
> remember that ontologies are the tail, not the dog, and the goal is to
> provide a way for software projects to produce better, more semantically
> accurate applications. It's the applications that typically bring
> business value though, not the ontology.
>
> Cheers,
> David
>
>> John
>>
>>
>>
>> _________________________________________________________________
>> Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
>> Subscribe/Config:
> http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
>> Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
>> Community Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
>> Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
>> Community Portal: http://ontolog.cim3.net/wiki/
>>
> (010)
--
Managing Director and Consultant
TopQuadrant Limited. Registered in England No. 05614307
UK +44 7788 561308
US +1 336-283-0606 (011)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
Community Portal: http://ontolog.cim3.net/wiki/ (012)
|