Re: Brainstorming Idea: Set up an online testbed
Leo,
That approach is reasonable, and I agree that focusing on
methodologies that have proved to be useful for practical
applications is extremely important.
As you know, I have also supported Robert Kent's IFF project,
and I have supported Philippe Martin's work on combining and
relating various ontologies. Their work is promising, and I
believe it should continue.
But I'd like to comment on the word 'community' in the following
message:
LO> However, the more times you go through the above process,
> and the more times you link two communities with different
> ontologies, the more you understand that commensurability/
> interoperability can best be solved by common ontologies,
> i.e., embedding those domain ontologies in a common ontology.
To make the discussion concrete, consider the Amazon.com
ontology, which they impose on all suppliers that hope to sell
products through Amazon (and that includes a very large number
of producers of books, electronics, etc.)
The communities that supply to Amazon are extremely diverse,
and there is little prospect of aligning all their ontologies
for a long time, if ever. But each of them can make their
computer systems interoperate with the Amazon computers by
aligning just a tiny subset of their ontologies with the
Amazon database schema.
There are many other database schemata that have been imposed
on diverse communities by various government agencies. Those
schemata have established de facto standards for interoperability
with current systems, and they will not disappear for a long time.
Furthermore, I strongly suspect that there is no such thing as
a true "community ontology", but rather that all ontologies are
collections of task-oriented specializations.
Bioinformatics, for example, is often considered a "community",
but if you look closer, it very quickly breaks down into a large
number of specialized subontologies, which are primarily task
oriented. The nurse's view is very different from the surgeon's
view, the pharmacist's view, the biologist's view, the accountant's
view, etc. The axioms for all those views are highly specialized.
What is common, however, are the terminologies. All those people
might use some of the same words to communicate effectively on
specific tasks, even though the detailed axioms they may associate
with the words have very little in common.
John