A working group for semantic interoperability
To the IEEE-SUO and CG study groups:
>From Patrick Cassidy (pcassidy at mitre dot org).
I am in the process of organizing a working group to try to develop
some level of coordination among the builders and users of Knowledge
Classification Systems (ontologies, taxonomies, thesauri and graphical
Knowledge Representations). The first meeting will be held on
Wednesday, October 5th, at MITRE facility in McLean Virginia, near
Washington. We plan to have a VNC connection, and a call-in line.
The charter for this WG and the general announcement are attached.
I am sending this note from my home computer (where I am
subscribed to the list) but communications will reach me faster at my
work e-mail: pcassidy at mitre.org (732-578-6340).
The working group is a subgroup of the federal Semantic
Interoperability Community of Practice (SICoP)
(http://colab.cim3.net/cgi-bin/wiki.pl?SICoP). The working group is
called tentatively the "Ontology and Taxonomy Coordinating Working
Group" (ONTACWG)
(http://colab.cim3.net/cgi-bin/wiki.pl?OntologyTaxonomyCoordinatingWG).
The effort is unfunded and will depend on voluntary contributions of
time from the participants. The SICoP is a working group of the US
government, but participation in this project is not restricted, and I
hope to get broad input from multiple communities, including foreign
groups, especially those in Europe.
Among other objectives, I hope that the ONTACWG can develop enough
agreement on the content of an upper ontology to provide a useful
conceptual "defining vocabulary" to allow specification of the meanings
of terms and concepts in any of the community knowledge
classifications, and in that way enable interoperability of
applications using inferencing. NOTE that "defining" is used in a
general, not technical sense here. Several purposes may be served by
a coordination effort, but as a minimum goal, I hope that this can
serve to increase the awareness among groups involved in knowledge
classification efforts of the specific aims and content of each other's
efforts, and how they relate to each other. If successful, this should
help reduce duplication and speed individual efforts.
I am hoping that efforts we can make within SICoP can complement other
efforts at relating knowledge bases to each other. To the extent
possible we will want to coordinate with efforts such as the NCOR
project that is being organized, and if possible, European efforts. As
a working group associated with the federal government, there may be an
initial focus within ONTACWG on various federal community knowledge
classifications. This should not inhibit participation of communities
with knowledge classifications oriented toward other domains. My hope
is for this working group to be able to make significant progress
toward eventual semantic interoperability. The benefits may initially
be mostly visible in improved browsing search, where taxonomies are
used to classify documents. Eventually, I hope that sufficient
precision in specifying the meanings of terms will be possible to
enable accurate inferencing.
It may seem a bit odd to some members of the IEEE-SUO group that, our
group having failed to agree on the relatively narrow topic of the
upper ontology, I would now propose to solve the problem by tackling
the larger issue of not only finding a common upper ontology, but also
mapping it to multiple domain ontologies and even to less formal
Knowledge Classifications Systems. What I hope is a legitimate method
in this madness is based on my belief that the main barriers to getting
agreement have been a lack of sufficient funding to allow the
participants to make commitments of significant time to the project,
and also the need within traditional standards organizations to seek a
high level of consensus even at the cost of long protracted discussion.
The ONTACWG initiative addresses the funding issue by assuming (after
ten years of trying) that the necessary level of funding will not be
obtained until a very broad range of builders and potential users of
Knowledge Classification Systems are convinced that the Common Upper
Ontology would address their needs; and will try to develop some such
conviction by showing that there is both a widespread desire to relate
KCSs to each other and a methodology to achieve that goal. This may
then convince potential funders that a Common Upper Ontology
developed by a group representative of diverse interests would be
widely used, and that developing one will not be a waste of money.
The problem of getting agreement may be soluble with three tactics; a
rapid decision-making process (voting, presumably) which is necessary
because of the very large number of issues that are involved in
building an upper ontology; an emphasis on finding a way to include
within an integrated conceptual structure as many as possible of the
alternative views on how to represent concepts; and development of
methods to selectively view portions of a large integrated ontology,
and to select out only those parts that are needed when the time comes
to build practical applications.
This committee provides a somewhat different mechanism from that used
within the IEEE-SUO effort to make progress toward wide adoption of a
common upper ontology. I do not expect that it will be possible, even
within a few years, for this committee to fully achieve development of
a complete upper ontology agreeable to all, nor complete mappings of
domain ontologies, nor development of utilities to make a common upper
ontology easy to understand and easy to use, unless significant funding
is eventually obtained. But I do think that we can make enough
progress to convince possible funders of the potential of this
approach, and in the interim to provide some valuable information to
the participants.
I hope that some members of the IEEE-SUO study group will participate
in both the construction of the Common Semantic Model (upper ontology
and extensions), and help in creating the definitions of some of the
domain concepts, using the CSM. There has been a lot of very useful
and informative discussion within this group, and I know that many
members could provide very valuable input to the ONTACWG project.
Pat
Patrick Cassidy
MITRE Corporation
260 Industrial Way
Eatontown, NJ 07724
Mail Stop: MNJE
Phone: 732-578-6340
Cell: 908-565-4053
Fax: 732-578-6012
Email: pcassidy at mitre.org
--
=============================================
Patrick Cassidy
MICRA, Inc. || (908) 561-3416
735 Belvidere Ave. || (908) 668-5252 (if no answer above)
Plainfield, NJ 07062-2054
internet: cassidy@micra.com
=============================================
ONTACWG.doc
GeneralAnnouncement.doc