[Top] [All Lists]

Re: [ontology-summit] Potential Tracks for Ontology Summit 2013

To: ontology-summit@xxxxxxxxxxxxxxxx
From: John F Sowa <sowa@xxxxxxxxxxx>
Date: Wed, 12 Dec 2012 17:20:37 -0500
Message-id: <50C90335.7080603@xxxxxxxxxxx>
Michael, David, and Fabian,    (01)

> ... you should look at the ontology-based infrastructure used by
> Mission Critical IT. Their platform is called ODASE.    (02)

Thanks for the reference.  But naming a platform "Mission Critical IT"
suggests a hope, but not necessarily a reality.    (03)

I'm happy to see that they integrate it with mainstream programming
languages such as Java and C#.  I am also happy that they are using
a logic-programming language (Mercury).    (04)

In their documentation, they did cite one actual application:    (05)

    http://www.missioncriticalit.com/pdfs/swese-2007.pdf    (06)

This was for an insurance company, which had been using "an internal 
domain specific language (DSL)."  I'm sure that a logic programming
language, such as Mercury, could support a language like DSL more
easily than Java or C#.  That language alone could account for any
advantage they have over more conventional programming.    (07)

Furthermore, this article does not say that the application was
mission critical or that it was actually deployed and used.    (08)

I repeat the question:  What mission-critical applications has anybody
developed with that platform?  Are they actually being used on a daily
basis?  For any significant time (at least a year) after the software
was delivered?    (09)

> I would disagree if anyone is suggesting that a good way to evaluate
> the quality of an ontology by merely whether it has been used in
> production systems.  Lots of crap is used in production.    (010)

I agree with the concluding sentence.  But I have even less faith
in the 99% of stuff on SourceForge that was developed by students
and never used by anybody except the developers.    (011)

It does not matter how elegant any software may be.  If it is not
deployed and used, it is a toy, an academic exercise, or a failure.    (012)

Academic exercises can be very instructive.  But they're not mainstream.    (013)

> Underspecification helps enable reuse, not necessarily interoperability.    (014)

First of all, I must emphasize the difference between specifying the
*data* at the interface, specifying the details of the operations
that each system performs on the data, and specifying the real-world
ontology of the entities and actions that produce or consume the data.    (015)

Independently developed systems can interoperate successfully when
they agree on the interfaces and some information about the data.
But their internal operations on the data may be very different.    (016)

As an example of a very successful and widely used ontology, I would
cite the Amazon.com DB schema.  Every business that wants to sell 
anything through Amazon (which includes a huge percentage of companies
that produce products for home consumers) must map their databases to
the Amazon schema.    (017)

The Amazon.com schema is very underspecified, but it enables thousands
of businesses around the world to interoperate successfully to sell
their wares through Amazon.    (018)

> In fact, in engineering ontologies, very specific constraints beyond
> what can be stated in logic languages are often required for
> interoperability to be possible.    (019)

When you require the full specification of the internals of a complex
application, you do need very detailed specifications.    (020)

But *every* constraint that can be implemented in software can be
stated in logic.  In fact, FOL can specify the details of every
digital computer and every program that can be implemented on it.    (021)

There are many (actually most) constraints that cannot be expressed
in OWL.  But that is a weakness of OWL, not of logic.    (022)

> there are many ontologies that are built without a specific system
> in mind and are used by different communities for different purposes;
> e.g., the Gene Ontology, the Foundational Model of Anatomy. The
> requirements for these reference ontologies cannot be derived from
> the requirements of some bigger IT system.    (023)

Yes.  These ontologies describe the data.  The systems that use them
require a great deal of detail about the structure of the data, but
the actions each system performs vary from one system to another.
Those actions are outside the Gene Ontology or the Anatomy Ontology.    (024)

And there is a huge amount of information about human biology that
is not in the gene ontology or the anatomy ontology.    (025)

John    (026)

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/OntologySummit2013/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2013  
Community Portal: http://ontolog.cim3.net/wiki/     (027)
<Prev in Thread] Current Thread [Next in Thread>