Dear Matthew, Pat, Doug, (01)
The following sentence is a succinct summary of all the issues: (02)
MW
> One of the problems with ontology design (as with most areas of design)
> is that there is more than one possible way to create something that works. (03)
My only qualification is that I don't consider it a *problem*. I
consider the open-ended number of options an opportunity that enables
engineers (including anybody who builds or modifies anything) to
adapt to the dynamically changing complexity of the world. (04)
PH
> There are certainly tricky edge cases. Carry a can of paint into a room
> of a house, paint the walls, let the paint dry. The dried paint is then,
> arguably, part-of the house. When did that happen? I honestly don't know,
> but it was somewhen in the above scenario. (05)
Very few people would ask or even think about that issue. They would
only think of asking if the answer were relevant to achieving some goal
or avoiding some hazard. (06)
At an early point in my career at IBM, I was in a mathematical analysis
group. People from other departments would come to us with questions
as puzzling as those. My first response was "Why?" (07)
After a few such questions, they would often realize that there was
a simpler way to restate their goals. My manager was not completely
happy with that solution -- because we lost the opportunity to bill
the other department for our services. (08)
DF
>> It seems that you would have to define interior boundaries to the
>> house so that you can exclude the contents of the house as not
>> being "part of" the house. (09)
PH
> Yes, exactly. I was presuming a topological notion of 'boundary'. (010)
For these and other issues, I would ask "Why?" (011)
In fact, one of my major criticisms of Cyc is that they tried to
anticipate and answer too many questions. But there is no limit
to what people can ask. For many questions, it's better to find
some way of restating the goals. (012)
That is one reason among many why I believe that the upper levels
should be as simple, general, and underspecified as possible. (013)
1. Every solution to every problem that has been solved should be
preserved and stored in a library of modules or microtheories. (014)
2. But it's impossible to anticipate and answer every question or
solve every problem. Recommendation: don't even try. (015)
3. Instead, design general tools that ask the equivalent of "Why?"
and help the users restate their questions and discover their
own answers as needed. (016)
In effect, those tools would play the role of Socrates in helping
the users develop or refine their own ontologies. As Kant said, (017)
IK (in the student notes labeled _Vienna Logic_)
> Socrates said he was the midwife to his listeners, i.e., he made
> them reflect better concerning that which they already knew, and
> become better conscious of it. If we always knew what we know,
> namely, in the use of certain words and concepts that are so subtle
> in application, we would be astonished at the treasures contained
> in our knowledge... (018)
And by the way, I would *not* require the users to learn any formal
logic or notation. Any system that can ask meaningful questions
must have some internal logic. It should be able to use that logic
to represent whatever answers the users discover. (019)
John (020)
_________________________________________________________________
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 (021)
|