[Top] [All Lists]

Re: [ontolog-forum] Relating and Reconciling Ontologies

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Sat, 23 Apr 2011 12:28:56 -0400
Message-id: <4DB2FE48.6010503@xxxxxxxxxxx>
Ali, Pat, Matthew, and Patrick,    (01)

> Cost in this context means that changing and admitting a novel
> core ontology should not be a trivial procedure.    (02)

The key term is 'admitting', and the next question is "Admitting
to what?"  To be concrete, suppose that somebody wants to submit
an ontology to some organization, say the OOR.  That organization
would establish various conditions.    (03)

For example, anybody could submit anything for free, and it
could be posted as an untested proposal.  If it fails various
tests, it would be rejected.  If it passes basic tests, it
could be posted as a proposal that has passed those tests.    (04)

Then other users could provide testimonials:  they tried it,
and it produced results that were (a) good, (b) bad, .....    (05)

A proposal that changes high-level assumptions might cause
considerable disruption, and many testimonials would say so.
But others might give glowing reviews that show how the new
ontology solves some critical problems.    (06)

That is the normal way that science and engineering progress.    (07)

> If the "purpose" is interoperability, then it does indeed make
> sense to have a single common foundation ontology.    (08)

But the word 'interoperability' has many different levels.    (09)

For example, the Amazon.com database interoperates with thousands
of DBs around the world.  But the amount of detail is limited:
product name, type, identifier (ISBN or part number), price, etc.    (010)

> The purpose of a common foundation ontology is to provide an
> interlingua for translation among a very large number of other
> ontologies; it is only needed for that purpose.    (011)

By that criterion, the database schema by Amazon.com is an
interlingua that enables interoperability with the ontologies
of thousands of DBs and KBs around the world.    (012)

That "bare bones" level of detail is typical of most interactions
on the WWW and the Semantic Web.  Key terms like 'product' are so
loosely defined that they cover anything from books, to cameras,
to groceries without giving any detail.    (013)

> The common foundation ontology can be of modest size - I suspect
> it can be kept to under ten thousand elements...
> But I am convinced that no other approach will work, and until
> the common foundation ontology is given a serious try...    (014)

> For interoperability it is not another foundation ontology that is
> needed, but a mapping between the different foundations that are already
> found. These in turn can be traced back to the ontological commitments made
> in those foundation ontologies, and how those ontological commitments
> translate in practical terms.    (015)

I agree.  But I would add that legacy systems are based on *implicit*
ontologies, which are just as important as any explicit ontology.
It's useful to make those ontologies explicit, but don't ignore
legacy systems just because their ontology is implicit.    (016)

Computers have been interoperating successfully since they were
first lashed together in the late 1950s.  Long before that, companies
interoperated by punched cards, banks interoperated by hand-written
notes since the 16th century, and don't forget cuneiform tablets for
international commerce in 3000 BC.    (017)

> You need perhaps a hundred or so concepts, so rather less than the
> 10,000 you mention.
> The other good thing is that there is nothing to mandate. The mappings
> deal with how different situations are dealt with under the different
> ontological commitments, and they can be developed from use cases that
> cover the usual difficulties. There are probably about 10-20 of those.    (018)

I agree that a small number of concepts is sufficient for interactions
at the level of detail from the the Sumerians to the Amazon DB schema.    (019)

Engineers need much more detail, but don't forget that in 1969, they
got hundreds of computers at companies and government agencies to
interoperate in putting a man on the moon.    (020)

> Curious if you have compared the growth of the Semantic Web with Esperanto?
> Thinking the higher % of semantics in a medium (think RDF/RDFa), the
> slower the growth rate (or even the possibility of growth).    (021)

It's not semantics that stopped the growth of Esperanto -- it's the
much more rapid growth of English after WW II.  English has the
largest semantics (in terms of the size of its dictionaries), but
it succeeded because of supporting infrastructure.    (022)

The adoption rate of RDFa has been rapid because it makes a simple
extension to HTML, it doesn't require a bloated notation, and it is
easy to integrate with all the programming languages and databases
used in commercial web sites.  (For the same reasons, PHP and AJAX
also grew much, much more rapidly than the SemWeb.)    (023)

And by the way, Esperantists ignored what I believe is the *only*
semi-artificial language that had any chance of succeeding: Peano's
Latina sine Flexione.  In 1903, it had a large infrastructure --
including a huge population that could learn it with much less
effort than learning English -- or even Esperanto.    (024)

I would compare the Esperantists to the Semantic Webbers.  They
tried to start from scratch and ignored the "low-hanging fruit".    (025)

John    (026)

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    (027)

<Prev in Thread] Current Thread [Next in Thread>