Dear Matthew, Ron, and Mike, (01)
Well-established groups have been developing domain-dependent
standards for a very long time. Our email forum can't compete
with them, and we can't expect them to replace their terminology
and definitions with ours. Instead, we must accept their documents
as standards and provide the means to accommodate and relate them. (02)
MW> What I am hoping for, is that for these basic things, the people
> responsible for their definition (the authoritative source) will
> provide suitable URIs. So for example, that the International
> Committee for Weights and Measures (Comité international des poids
> et measures) will provide the ones for units of measure. I think
> that an authoritative source can be identified for most of the
> things you mention above. The problem I see is in persuading them
> to rise to the challenge of providing the web identities and
> definitions. Perhaps we could help? (03)
We could help by providing an easy-to-use registry that generates
unique identifiers. For example, nearly all businesses in the US
have DUNS numbers issued for free, upon request by Dun and Bradstreet.
They ask for and record some basic information, but they make the
process as painless as possible. As a result, they have become the
de facto standard for providing unique identifiers that are widely
used in the databases of businesses and governments. (04)
Definitions, however, are a very different matter. We cannot force
anyone to accept our definitions, but we can record and point to
theirs. If two different groups X1 and X2 define the same term T,
we must distinguish them as T.X1 and T.X2. Some other group X3
might claim that T.X1 and T.X2 are identical or not identical.
If so, we record that fact. The ontologies and definitions that
are highly recommended will be widely used, and the ones that are
ignored will sink out of sight (while remaining in the archive for
historical purposes). (05)
RW> The marketplace may be cruel but it does eventually pick a set
> of winners. Big players such as IBM, US government, DOD and
> Microsoft have the power to distort the process but that is just
> a fact of life. (06)
I agree, and I believe that a good registry can aid fairness and
promote the selection of the best versions. By making *all*
versions and *all* reviews of those versions easily available,
the users (AKA "the market") will have good information for
making their decisions. (07)
MB> There have also been turf wars of course. (08)
And we should stay as neutral as possible, not take sides,
or worse, create more fighting. (09)
MB> Meanwhile it should be noted that ISO has a very good process
> for taking an existing stable standard from one country or
> organisation, and anointing it as an international standard.
> What it does not have, however, is any good business process
> for developing anything itself. (010)
I first became involved with ANSI and ISO in 1991, when there were
many people who wanted those groups to become more "proactive" in
developing standards from scratch. But their best work has always
been to clean up, polish up, and harmonize de facto standards. (011)
The Common Logic standard, for example, is based on harmonizing
two independently developed versions of logic that had long-term
communities of users: conceptual graphs and KIF. In the process,
they also "webized" both of those languages by adding support for
Unicode and URIs and making them upward compatible with RDF and OWL.
The result, I believe, was better than either of the two starting
languages. (012)
The best W3C "recommendations" (the W3C is not a standards body)
are the ones based on languages and interfaces that were independently
developed, implemented, and used. I believe the major weaknesses of
the Semantic Web languages were caused by the W3C being "proactive":
they recommended RDF and OWL long *before* there was any significant
usage. As a result, the languages were frozen in something like
version 0.2. A lot more has been added, but they propagated the
mistakes of version 0.2 instead of removing them. (013)
MB> I do think that the ontology standards need to go beyond "this
> is a Thing" and implement some common sense rules for how context
> and other aids to semantics can be made to work. (014)
That's a good area where we can be proactive. Commonsense rules and
guidelines don't get frozen in code. (015)
John (016)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (017)
|