At 11:25 PM -0400 4/26/08, John F. Sowa wrote:
>Pat,
>
>There is no difference in principle between compiling a
>procedural representation (such as C) or a declarative
>representation (such as OWL or predicate calculus) to some
>internal form that is extremely difficult to reverse engineer. (01)
Well, I think there is a real difference, in that there is no
coherent accepted notion of 'compile' for the declarative case as
there is for the procedural one. But lets not quarrel about that. My
main point was only that the version in the OOR is going to be the
pre-compiled one, which is the more humanly readable one. (02)
>
>> How? You are saying that an ontology can be usable as an ontology
>> but cannot be reverse engineered to be humanly readable?
>
>For example, one could translate an OWL ontology to Prolog code,
>which is humanly readable. (03)
OWL is humanly readable, probably more so than Prolog (since it is
state-free). But in any case, I don't think one can translate OWL to
Prolog without seriously compromising the OWL semantics. (04)
> But then it is possible to compile
>that Prolog code to a C program and from C to machine code. (05)
Something is seriously wrong here. Machine code and OWL are
incommensurate. The OWL spec is careful to not give any conditions on
an OWL processing engine: the only semantics in the specification is
model-theoretic. Machine code (or C) has no model theory whatever,
and its semantics are concerned only with state-transitions (or
continuations, if you prefer.) Somewhere in this series of
'compilations' you have abandoned the OWL semantics and replaced it
with a procedural specification of some kind. (06)
> A
>reverse compiler from machine code might be able to generate
>a C program without comments, but it would be extremely difficult
>to reverse engineer the original Prolog or OWL from that C code.
>
>JFS>> If those internal forms are kept proprietary, the effect
>>> would be to keep the ontology as "closed" as any program.
>
>PH> Except that the source is freely available in the OOR, right?
>
>Yes. That's why it's called "open". (07)
Which was my whole point from the start of this thread. All this
misleading comparison to programming language compilation is beside
the point. (08)
> Many businesses want to
>provide services to their customers, but they do not want their
>competitors to copy the ontologies that drive those services.
>For that reason, they are not likely to deposit their operational
>ontologies in the OOR. (09)
So far, most of the application of ontologies I have seen has more to
do with interoperation and intercommunication than with the driving
of proprietary applications. Businesses which want to communicate
with their customers might well want to publish an ontology intended
to be used to facilitate this communication. (010)
Pat (011)
>
>John (012)
--
---------------------------------------------------------------------
IHMC (850)434 8903 or (650)494 3973 home
40 South Alcaniz St. (850)202 4416 office
Pensacola (850)202 4440 fax
FL 32502 (850)291 0667 cell
http://www.ihmc.us/users/phayes phayesAT-SIGNihmc.us
http://www.flickr.com/pathayes/collections (013)
_________________________________________________________________
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 (014)
|