On Tue, December 18, 2012 10:53, John F Sowa wrote:
> Dear Matthew, Pat, and David, (01)
> JFS
>>> As I said in my previous note, a formal ontology is stated in logic.
>>> That logic embodies the total meaning of that ontology. (02)
> MS
>> That is not really true (as you know). The model theory also has a part
>> to play, i.e. the intended interpretation. (03)
> I agree that the *intended* model is critical to the way human users
> and developers connect the computer and its software to the world. (04)
> But the computer has no access to our intentions. Nothing that we
> put in the comment fields of an ontology will have the slightest
> influence on what the computer does. (05)
Yes, John. But the human users and developers do not just give an
ontology to a logic engine and let it have its merry way. They write
programs to interact with the real world that use information encoded in
the ontology to make changes in the real world -- either physical changes
(devices being turned on or off or operating with different settings) or
changes in intangible
objects (amounts being added or deducted from accounts). The programmers
use the comment fields (and often term names) in order to make such
connections between the ontologies and the real world. (06)
Of course, those programs have a great influence on what a computer does
(if the program is run). In fact, such programs affect the actual (as
opposed to intended) meaning of the terms in the ontology. (07)
> MS
>> It comes as a surprise to some that the text definitions we give objects
>> in an ontology are not part of the formal semantics (what can be
>> interpreted
>> by computers) but is actually part of the model theory, and tells you
>> what
>> these things were intended to represent. (08)
> Human intentions are not involved in the model *theory*. But they
> determine what kind of data people feed the computer. If they feed it
> garbage, the computer will populate the model with garbage. Then it
> will process that garbage to generate more garbage: the GIGO principle. (09)
> To be precise: The logic of an ontology completely determines the
> meaning for the *computer*. (010)
Actually, the logic of *programs* that use the ontology determines the
meaning for the *computer*, not just the logic of the ontology. Such
programs are not restricted to first-order or sub-first-order inference
engines. (011)
> The comments in the logic may state
> the human intentions. But they are not part of the formal ontology,
> and they have no effect on what the computer does with the data. (012)
I agree that they have no *direct* effect on what the computer does
with the data. However, the programs written by developers who rely
on those comments *do* affect what the computer does with the data. (013)
> ... (014)
> But as Matthew said, the *intended model* is critical for any useful
> application. I agree. (015)
> But those intentions are enforced by data entry clerks or people who
> click on an iPhone app. Their way of thinking and talking is also
> disjoint from the terms being debated on this list.
> ...
> Summary: For specifying an ontology, our metalevel terms should have
> a simple mapping to whatever logic we use. For the *intended* models,
> some software designer must relate the comment fields of the ontology
> to the terms of the end user -- but that is a very different mapping. (016)
Agreed. (017)
-- doug f (018)
> John (019)
_________________________________________________________________
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 (020)
|