On 12/23/2010 2:36 PM, Christopher Menzel wrote:
> I think there is still a pretty pervasive idea (despite regular
> rejoinders from the likes of John Sowa, Michael Grüninger,
> Pat Hayes, Leo Obrst, etc) that building quality ontologies
> doesn't require any sort of special technical training, just
> a clear head, a bit of common sense, and a copy of Protege
> (an excellent and useful tool, let me hasten to add). (01)
Please do *not* attribute those claims to me. I definitely
do *not* believe that the average college graduate is capable
of defining any useful formal ontology. (02)
The reason why I have been promoting controlled NLs is that
I have abandoned all hope of getting them to use any formal
notation with any degree of accuracy. At best, I would hope
that they might be able to read CNLs, but even that is a stretch. (03)
> To the contrary, at a minimum, a competent, well-educated
> ontological engineer must have a mastery of first-order
> logic and related systems like modal logic (of which
> description logic can be considered a variety) as well
> as a deep familiarity with the history of knowledge
> representation and KR systems. Broad knowledge of the
> history of philosophy as well as contemporary metaphysics
> and philosophy of language would not only sharpen analytical
> skills and provide a rich source of ontological exploration
> to draw upon, but would also prevent ontological engineers
> from wasting time reinventing the wheel (typically with
> an inferior design). (04)
I agree with that list of requirements. I would also add
training in at least one language other than their native
language, and some study of linguistics and lexicography. (05)
Please note that the field of terminology is much simpler
than formal ontology, but professional *terminologists*
are expected to have extensive of training + experience
beyond a bachelor's degree. (06)
> Until ontological engineers, like engineers of every other
> other stripe, can be assumed to have a well-defined baseline
> of knowledge and a basic technical skills, an endless repetition
> of elementary modeling errors and, consequently, a stream of
> (at best) unreliable and (at worst) incoherent ontologies are
> to be expected, and skepticism about the usefulness of ontologies
> will (justifiably) persist. (07)
I agree, but I don't expect that to happen for a long time to come.
That is why I believe that the attempts to get people to add
semantic annotations to documents is a dead end -- there just
aren't enough trained humans to do that kind of task with any
degree of reliability. (08)
Instead, I believe that the NLP technology is increasing at
a sufficiently rapid rate to support automated or at least
semi-automated tools. (09)
And I do not expect the humans who use the semi-automated
tools to have any formal training in logic, linguistics,
or ontology. Instead, I would only expect them to be able
to tell the difference between plausible and nonsensical
statements in their own area of expertise. (010)
John (011)
_________________________________________________________________
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
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (012)
|