Pat, (01)
That is a much more restricted and a much more achievable
goal than a common upper ontology. For most of those
things, there already are official standards or de facto
standards: (02)
1) The Types for "Named entities"... (03)
person names
person titles
company/organization names
geopolitical entity names
locations and geographical places
dates×
individual events (e.g. World War II)
percentages
monetary amounts
measures (04)
The terminologists, the ISO standards groups, and various
professional societies agreed to standards in these areas
many years ago, and in some cases, centuries ago (e.g., biological
species, astronomical objects, chemical elements and compounds,
UN and EU organizations, etc.). There are even ISO standards
for screw threads and grades of wheat. (05)
I agree with John Bateman that people have given up waiting
for a common ontology and with Mike Uschold that computer
systems have been interoperating on shared databases since
the late 1960s without having a common ontology. What they
have been using are shared terminologies, such as the above.
That is where "the action is" right now. If we want to be
taken seriously, we have to build on those standards. (06)
New systems must always coexist with running systems, and
large systems have lifetimes of 20 to 40 years. Without
a migration path from legacy systems and a coexistence
strategy for current systems, the best defined ontology
will remain an academic exercise. (07)
And I'd like to thank Mike for the following point: (08)
> I'm expecting John Sowa to chip in here, but I will save him
> the trouble...
>
> So in what sense is RDFS more advanced than RDBs/SQL? (09)
I would concede one point: the typed version of RDF does
recognize types, which Ted Codd proposed about 30 years ago
for RDBs, but which still aren't in the SQL standard. But
for performance, reliability, security, etc., RDF & OWL are
still infants compared to RDBs, which run the world economy,
both now and for a long time to come. (010)
Suggestion: One of the first tasks we should do is to compile
an inventory with pointers to all the official and de facto
standards for the named entities that Pat listed above.
Any proposed ontology must accommodate all of them. (011)
John Sowa (012)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uos-convene/
To Post: mailto:uos-convene@xxxxxxxxxxxxxxxx
Community Portal: http://ontolog.cim3.net/
Shared Files: http://ontolog.cim3.net/file/work/UpperOntologySummit/uos-convene/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UpperOntologySummit (013)
|