John here describes precisely the goals of the Bioportal as
conceived by the National Center for Biomedical Ontology
(http://ncbo.us) for the
domain of biomedical ontologies (but including also those upper level
ontologies, such as Cyc, SUMO, Dolce and BFO, which have been used in
this domain).
BS
At 05:35 AM 3/15/2006, you wrote:
Adam,
The option of having a collection of modules (call it a
library, registry, lattice or whatever) does *not* preclude
the option of having preassembled versions, such as Cyc,
SUMO, Dolce, etc.
AP> In fact, the characterization of "monolithic"
ontologies
> seems to me just pejorative since SUMO, Cyc and DOLCE all
> have their different approaches to modularity.
Fine. We can call it whatever you like, but we should
document all the choices, put them in a registry, and
let the users decide.
AP> I don't think we need building blocks, but rather an
> integrated and common standard. Building blocks require
> further assembly.
We can have *both*. Right now, we have several
"candidates",
and none of the developers have shown the slightest inclination
to adopt any of the others in preference to their own.
If we have a registry of ontologies, we can include big ones
and little ones -- tiny modules and larger composites. For
each one, the metadata can show all pertinent information that
may help users decide which to choose:
1. Who defined, tested, and used each one. What is its
complete development history with full
documentation.
2. What modules, if any, were included in it, and which
composites contain it as one component.
3. How is it related to other alternatives, as a
generalization,
specialization, relabeling, or combination of some
other(s).
4. If it is incompatible with other options, what are the
advantages and disadvantages of the
alternatives. Under
what conditions might one prefer one or the
other.
In effect, the registry becomes a complete catalog of all
ontologies -- big composites and smaller modules -- with
all relevant documentation, including the complete history
of development and use. Any user can add metadata to the
registry to document experience, positive and negative, with
any of the big composites or smaller building blocks.
Instead of endless arguments, we can make all options available
and let the users decide. If there is an advantage to using
a particular ontology, the users will gravitate to it and will
leave an audit trail of their choices and experiences.
John Sowa
_________________________________________________________________
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
|