Dear Steve, (01)
Responding to your "opening remarks" below. (02)
Regards (03)
Matthew West
Reference Data Architecture and Standards Manager
Shell International Petroleum Company Limited
Shell Centre, London SE1 7NA, United Kingdom (04)
Tel: +44 20 7934 4490 Mobile: +44 7796 336538
Email: matthew.west@xxxxxxxxx
http://www.shell.com
http://www.matthew-west.org.uk/ (05)
> Upper Ontology Summit Panelists and Key Participants:
>
> This note provides additional detail about the coming
> Upper Ontology
> Summit in Gaithersburg, MD, and its rationale. The goals and
> approach will
> ultimately be that decided by the eight invited panelists.
> The organizing
> committee have prepared some starting suggestions.
>
> The two broad goals proposed by the organizing committee are: to find
> methods to interrelate existing upper ontologies so as to
> permit users of
> each to reuse knowledge expressed in the others; (06)
MW: I think this is an excellent idea. (07)
> and to
> increase public
> awareness of the maturity and capabilities of ontology technology for
> automating information analysis and exploitation. (08)
MW: On the one hand I think that considerable progress is evident. On
the other hand I think we are a long way from anything that could really
be described as "mature". I recommend cautious optimism. Also be prepared
to defend the current initiative as an act of desparation. (09)
> Correspondingly, the discussions leading up to the meeting
> and the meeting
> itself will deal with both technical and public-relations
> issues. The goal
> of improving the public perception of the current state and
> potential for
> ontology technologies will be as significant as the technical
> questions to
> be explored.
>
> With respect to the proposals for methods to relate the upper
> ontologies to
> each other, it is understood that substantive efforts would
> require some new
> funding specifically for that purpose. (010)
MW: Certainly true. (011)
> We are hopeful that positive
> publicity coming from this Upper Ontology Summit could
> improve the public
> perception of the state and importance of formal ontology
> development, and
> thereby increase the chances that funding could be obtained
> specifically for
> that effort at interrelating the ontologies, as well as for
> both research on
> and commercial development of ontologies related to one or
> more of the upper
> ontologies you represent. (012)
MW: Although funding for direct development of ISO 15926 would
always be welcome, we do have at least a modest stream of
funding available for this - driven by industry need.
We have no funding for inter-relating ontologies beyond some
limited volunteer effort. (013)
> We hope that the public communique
> will express
> at least an agreement in principle to work toward some means
> of relating the
> ontologies to each other. (014)
MW: I'm certainly prepared to sign up for that (subject to
funding).
>
> This note describes:
> (1) suggested methods for discussion among the UOS panelists
> prior to the
> meetings March 14th and 15th;
> (2) The schedule for March 14th and 15th.
> (3) A specific proposal for one method to relate the upper
> ontologies to
> each other; this can be discussed and modified, or ignored in favor of
> alternative ideas presented by the panelists;
> (4) Segments of a sample communique that might be agreed to and made
> public at the conclusion of the March 15th meeting. Any part of the
> sample may be adopted, changed, or discarded by the panelists. Other
> agreements reached that the panelists would like to announce
> publicly would
> be additionally helpful for creating a positive impression of
> the maturity
> and potential of the field.
>
> In addition to the panelists and organizers, certain "key
> participants" are
> hereby cordially invited to participate in the discussion
> prior to and at
> these meetings. Those key participants are :
>
> Bill Andersen
> Dieter Fensel
> Chris Menzel
> Mark Musen
> Jim Schoening
> Mike Uschold
> Chris Welty
>
>
> On behalf of the organizing committee, and also from a
> personal perspective,
> I sincerely hope that I will be seeing you here in
> Gaithersburg in March. (015)
MW: My flights and hotel are booked. (016)
<snip>
> (3) One possible goal to be discussed: A Common Subset
> ontology Compatible
> with each of the Upper Ontologies
>
> On the Upper Ontology Summit page:
> http://ontolog.cim3.net/cgi-bin/wiki.pl?UpperOntologySummit
> . . . there is a straw-man starter proposal to attempt to
> build a "Common
> Subset Ontology", with a structure translatable into each of
> the linked
> fully-axiomatized upper ontologies.
>
> http://ontolog.cim3.net/cgi-bin/wiki.pl?UpperOntologySummit#nidJF4
>
> The rationale for this proposal is that a simplified less
> highly axiomatized
> ontology that is compatible with all of the linked upper
> ontologies can help
> to find relations among the upper ontologies while
> simultaneously providing
> guidance to the large community of potential users of upper
> ontologies. With
> sufficient axiomatization to provide precise specifications
> for the common
> concepts of person, organization, time, event, etc., the common subset
> ontology can help to introduce those already building simple
> OWL ontologies
> to the need for and possibility of precise logical definitions of the
> semantic relations.
> By encouraging use and testing of a simplified upper
> ontology, demand for
> the more complete and more functional upper ontologies should
> follow as a
> consequence of improved understanding of the potential.
> Allowing users to test their own systems with a simplified
> ontology that can
> be converted into any of the more highly axiomatized
> ontologies, will reduce
> or eliminate the hesitation to spend resources on a
> technology that may
> become rapidly obsolete. What is built with the compatible subset
> ontologies will be reusable with each of the linked upper ontologies
>
> The panelists will have the opportunity to discuss the
> proposal for a common
> subset ontology and other methods for relating their
> ontologies to each
> other. (017)
MW: I think the idea is good in principle, but I think needs to be set
in a slightly different direction. The current direction seems to be
one of a merged lowest common denominator approach. I think that to be
successful, we will have to move to a lattice of theories approach, as
described by John Sowa. Here, basic theories are stated separately,
and can be combined in a mix and match approach. In this I would see
in particular the possibility of sharing a hierarchy of natural and
intentional kinds, independent of whether these were thought to be
intensional or extensional and whether instances were 3D or 4D
as perhaps the highest value activity that could be undertaken.
>
> (4) Sample of a possible Communique to be announced on March
> 15th as a joint
> statement by the panelists.
>
> =============================
> Sample Communique
> =============================
>
> Joint Statement of the Panelists of the Upper Ontology Summit
> March 15th,
> 2006
>
>
> Background
>
> The science and technology of representing knowledge in a
> form suitable for
> use and reasoning in computers has been developing for over
> thirty years.
> Practitioners in the field have achieved an increasingly detailed
> understanding of the fundamental components of meaning and
> how to represent
> them in formats suitable for computer processing. With the
> success and
> expansion of the internet, the potential for creation of a
> Semantic Web has
> become widely recognized, and the number of teams and
> individuals creating
> knowledge classifications of varying degrees of logical formality has
> dramatically increased. As this technology develops further
> it will enable
> deployment of computer applications with increasing ability
> to make reliable
> knowledge-based decisions that currently require human
> effort. Programs
> with such enhanced capacity will increase the speed and efficiency of
> automated information analysis and exploitation.
>
> Much recent emphasis has been focused on creating common syntactical
> formalisms for representing knowledge, but syntactical
> formalisms alone
> do not provide an effective standard of meaning. The complementary
> technology for effectively representing the semantic content
> of complex
> widely used concepts is also available, but agreement on a
> standardized set
> of conceptual elements has not yet been reached. The need for such
> agreement is increasing rapidly as many isolated projects of varying
> complexity have been initiated to capture knowledge in
> computer-interpretable formalisms. Without a common resource for
> specifying meaning in a uniform fashion, the great potential
> for sharing
> knowledge usable for computer reasoning will not be realized.
> A new initiative is needed to find agreement on a set of
> basic concepts that
> can be easily understood and exploited by diverse communities
> of users to
> permit their systems to accurately exchange detailed meanings for the
> concepts they need to communicate with each other. (018)
MW: This rather glosses over the 3D/4D divide, almost certainly
appropriately for our audience. Just as long as we do not kid ourselves.
>
> The basis for detailed representation of meaning is now
> available in several
> upper ontologies together with associated mid-level and
> domain ontologies,
> reflecting the results of decades of research and development
> in knowledge
> representation. Each of these ontologies has an existing community of
> users. Finding a means to relate these ontologies to each
> other and to make
> them easily usable by other developers of domain ontologies
> can provide
> users with access to the essential common standard of meaning
> that will
> allow accurate interchange of conceptual information among multiple
> communities, significantly enhancing the value of the
> knowledge in each of
> the communities whose knowledge bases are linked to the
> common standard. (019)
MW: OK.
>
> Conclusion of the Upper Ontology Summit
>
> The theory and technology of knowledge representation have
> advanced to a
> stage where the concepts that are the meanings of terms can
> be formally (020)
MW: Ontology is supposed to be about what exists, not the meaning
of terms. (021)
> specified in computer systems with great detail and precision.
> To demonstrate the power of the technology and provide
> benchmarks by which
> objective evaluations of alternative methods of representing
> knowledge can
> be tested, it will be helpful to create sophisticated
> open-source applications available for public evaluation. Each of
> the existing upper ontologies can provide a basis for a
> standard of meaning
> representation, but the details of the representations are
> sufficiently
> different that knowledge in one ontology cannot at present be
> accurately
> translated into another. (022)
MW: I think the real issue is one of expense. I am confident accuracy
can be achieved if enough money was thrown at it (applied through
sufficient brain power). (023)
> To ease the creation of
> applications that can take
> advantage of any of the existing upper ontologies, (024)
MW: My motivation would be as much to seek improvements to the
individual ontologies that would remove/reduce unnecessary differences
between them. I've seen this work with data models. (025)
> we now
> want to develop
> methods to relate the existing upper ontologies to each other. (026)
MW: I think this is the right approach (with some shareable elements
perhaps resulting from this, rather than being created by some sort
of merging effort).
>
> We have agreed to take substantive steps toward developing methods to
> interrelate the existing upper ontologies to each other. The result
> of a successful effort will provide a standard of meaning that will be
> freely available and readily adapted as a basis for any
> domain ontology. It
> will enable a high level of interoperability, and a means to
> develop basic
> ontologies that can be incorporated into more functional
> ontologies using
> any of those upper ontologies linked to it.
>
> Conclusion of the Upper Ontology Summit in Brief
>
> (1) We agree that the technology of modeling and representing
> knowledge has
> developed to the point where it is feasible to create knowledge-based
> reasoning systems with information analysis and exploitation
> capabilities
> significantly more advanced than traditional systems based on
> relational
> databases and object-oriented programming without semantic
> interpretation.
>
> (2) Each of the existing upper ontologies differs in specifics of
> implementation, but we all agree that use of some formally
> defined common
> upper ontology is essential for semantic interoperability. (027)
MW: What would be common would hardly be itself a common upper ontology,
(as I see it) but rather some elements that could be common to all the
upper ontologies.
>
> (3) Each of the existing upper ontologies has different
> strengths, and we
> believe that the technology will be advanced most rapidly by
> continuing
> exploration of such different approaches, while finding the
> common elements (028)
MW: This is more what I think we should be doing. (029)
> that will help users develop applications that can take
> advantage of any of
> the well-structured formal upper ontologies. (030)
MW: Well I would hope that improvement of the individual ontologies
through exposure to the ideas in others would tend to lead to a merging
of many of them as improvement individually reduces the difference between
them. But this should not be forced.
>
> (4) We believe that finding methods to relate different
> upper ontologies to
> each other will have near-term practical benefits in enhancing
> interoperability of knowledge-based systems, and will also permit more
> effective investigation of reasoning methods on large knowledge bases. (031)
MW: Do you have some examples where interoperability is required, but
not achieved? It would be useful to quote them.
>
> (5) As one means of interrelating the existing upper
> ontologies, we plan to
> explore the possibility of creating a common subset ontology (032)
MW: I would prefer the common elements that could be shared language,
rather than setting up the effort as producing another ontology. (033)
> that will be
> accurately translatable into each of the linked upper ontologies. In
> addition to making the existing upper ontologies more easily
> accessible,
> this can provide a simple but powerful tool to allow users to explore
> creation of interoperable systems with basic reasoning
> capability on a pilot
> scale.
>
>
> (034)
_________________________________________________________________
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 (035)
|