RE: Call for vote on SUO Scope and Purpose
Matthew,
Thanks for the clarification about standard evolution. What you suggest
makes sense. Would you be willing to draft a paragraph explaining a
process for SUO evolution which could be merged into a future version of
the Scope and Purpose?
Adam
At 10:40 AM 7/27/2000 +0200, West, Matthew MR SSI-GPEA-UK wrote:
>Dear Colleagues,
>
>Bill requested
>
> >>(5) The SUO should have some provisions for the evolution of the SUO over
>time.
>
>Adam responded
>
> >That seems to be more an issue for the standardization process than for the
>standard itself.
>
>I can assure you that if you leave this to the standardisation process you
>will have a torrid time. Imagine that you have developed a standard upper
>level ontology, and you find there is a mistake or an omission. If you rely
>on the standardisation process, this can take up to 5 years to include or
>correct.
>
>Both Bill and I are part of the ISO TC184/SC4 - Industrial Data - community,
>and we have both suffered from precisely these problems.
>
>One approach we are beginning to adopt is somewhat consistent with what
>Robert Kent was suggesting. This is the use of a Register. Here you have an
>ontology (or schema) for an ontology, which includes those things necessary
>to manage additions and change (maintaining a history of what the ontology
>was historically, and supporting a change and approval process).
>
>Two standards in ISO TC184/SC4 are being developed following this approach.
>
>The first is ISO 15926 - integration of life-cycle data for Process Plant
>(including oil and gas) facilities. (Note: don't be fooled by the title, as
>someone once remarked "You can even use this to model Wellington Zoo")
>
>The second is ISO18876 - Integration of Industrial Data for Exchange Access
>and Sharing (IIDEAS). This aims, rather than supporting a single ontology
>(or schema) to provide facilities to integrate and support multiple
>ontologies/schemas, with the objective of being able to migrate information
>according to one ontology/schema to another where it is relevant.
>
>Regards
> Matthew
>============================================
>Matthew West
>Asset Information Management
>Shell Services International
>H3229, Shell Centre, London, SE1 7NA, UK.
>Tel: +44 207 934 4490 Fax: 7929
>E-mail: Matthew.R.West@is.shell.com
>http://www.shellservices.com/
>============================================
>
>-----Original Message-----
>From: apease [mailto:apease@teknowledge.com]
>Sent: 27 July 2000 01:49
>To: WBurkett@pdit.com; 'Schoening, James R CECOM DCSC4I'
>Cc: SUO (E-mail)
>Subject: RE: Call for vote on SUO Scope and Purpose
>
>
>Bill,
> I'm puzzled by some of your comments. Comments below:
>
>At 05:34 PM 7/26/2000 -0700, WBurkett@pdit.com wrote:
>
>
>James:
>
>I have reviewed the scope and purpose described below, and following these
>discussions as closely as my background knowledge allows. From these, I
>believe that I must vote "No with comments". This is not because I disagree
>with the intent of the work, but because I believe there are many very
>important things that are omitted from the scope/purpose statement. Before
>I get to those, it appears de riguer to provide an explanation of my
>background to provide a background context and perspective for my comments.
>
>Background: I am systems engineer by education and practice and have largely
>worked in the automation systems divisions for large manufacturing
>organizations (either directly or as a consultant); my particular area of
>focus as been on integration and application interoperability. As part of
>this work, I've been developing data models and participating in the
>development of international data exchange standards for (almost) the past
>20 years. I have also been pursuing university research on the question
>"what makes a data model good?", which has lead me into many decidedly
>non-engineering fields such as philosophy, linguistics, and even sociology.
>(Directions that have been reaffirmed by the discussions on this list.)
>
>Perspective: This background has lead me to a perspective on the issue of
>application system integration/interoperability that includes the following
>beliefs.
>
>* I equate "ontology" with "schema".
>* I believe that "meaning" and "knowledge" only exist in the human mind.
>Every externalization of this "stuff" is merely a representation that has no
>inherent meaning. (A corollary to this is that I believe that that
>anthropomorphizing of computer systems is a self-defeating mistake because
>by confusing the roles of machine and human it interferes with clear
>understanding of what computer systems can and can't do. Terms like
>"Inference engines" and "knowledge representation" refer to clever data
>processing to my mind (and my apologies to those whom this statement might
>offend :-) .)
>
>* I believe that whatever-it-is that happens in a human mind that is
>"meaning" is a real-world phenomena that should be part of the world view
>that backs the SUO we develop. (Thus, "possible worlds" do exist in the
>real world as mental phenomena.) Said another way, I think it is a big
>mistake to omit the human mental element from our account of "meaning" in
>the SUO because that is the only place in the universe where "meaning"
>really means anything. (Because if human minds didn't exist, the question
>of whether or not there is any "meaning" out there in the world is
>completely moot.)
>
>* Any physical manifestion of meaning - i.e., physical representations
>intended to recreate the a selected phenomenon in my mind within the mind of
>another person perceiving the representation - is at best an imperfect and
>imprecise mechanism.
>
>
>Given this background and perspective, my comments on the scope and purpose
>are as follows:
>
>(1) I believe that the objective of 1000-2500 terms is far too large to be
>practical. In my experience, I've found that the practical upper limit on
>the number of independent object/entity types in a schema is ~300 types.
>Models larger than this simply cannot be comprehensively understood by a
>single person. (One may argue that it's possible to understand larger
>models with the aid of a meta-level organizing structure - and I agree - but
>then *this* structure becomes the upper level ontology.) (And there is the
>further question of criteria for selecting and discriminating between
>ontological elements.)
>
>
>Folks in the HPKB project including Cycorp, Teknowledge and Stanford used
>the Cyc upper model of 3000 terms successfully in several tests. I take
>this as an existence proof that refutes your assertion, or more gently,
>maybe it just refutes the need for an individual to comprehend the entirety
>of a model for that model to be useful.
>
>
>(2) I think the SUO should be specified in a formal language to facilitate
>data processing. I don't think terms and definitions are enough for
>practical applications.
>
>
>Specification in a formal language is precisely what is intended "..An
>ontology is a set of terms and formal definitions..." I've proposed KIF or
>some derivative thereof which is a formal language.
>
>
>
>(3) One of the most important pieces missing from the Scope/Purpose is any
>recognition of the need for practical guidance for *using* the ontology.
>It's nice to say "If system developer use the SUO for the basis of their
>ontology, then there will be some degree of interoperability", but without
>some practical guidance on how this can be made to true, it is just hot air.
>*How* should the SOU be used?
>
>
>I think it's sufficient to say *what* the SUO might be used for. To specify
>how it might be used in any application would take textbook articles or
>tutorials. We might refer people to Russel and Norvig's AI text which has a
>good section on use of ontologies. What more do you believe is needed?
>
>
>(4) As I stated above, I think it is a big mistake to divorce the SUO from
>human minds. Do so would be similar to assumed objectivity of physical
>sciences (i.e., independence of natural phenomena from observer) that was
>discredited by, for example, Heisenberg's Uncertainty Principle (the
>observer affects the thing observed). Our goal in the specification of the
>meaning of the components of the SUO must include the recognition that a
>precise definition that means the same thing to all people at all times is
>impossible and seek instead to simply try and maximize it.
>
>
>I agree with this. Do you feel we need a disclaimer in the SUO to this
>effect or something more?
>
>
>(5) The SUO should have some provisions for the evolution of the SUO over
>time.
>
>
>That seems to be more an issue for the standardization process than for the
>standard itself.
>
>
>(6) Although I recognize that we don't really have any other choice, I am
>leery about using English to define/specify the ontology. The english
>language has its own ontology, but is there anything that makes English a
>better specification language than, say, Italian, other than its ubiquity?
>Is the bootstrap nature of the definitions going to be a problem? I don't
>know ...
>
>
>The SUO Scope and Purpose does not in any way state that English will be
>used to define the ontology.
>
>
>Summary: On the whole I applaud the objectives of this group and following
>the discussions with great interest (and some degree of understanding ;-) I
>hope that the ground rules and objectives can be solidified to a degree that
>the outcome will be of practical use in system integration and
>interoperability.
>
>
>I'm excited about the group too and hoping that many of your concerns rest
>on some misunderstandings which hopefully I've been able to clarify.
>
>Adam
>
>
>Bill Burkett
>
>
>
>------------------------------------------------------------------
>William C. Burkett 562-495-6500x13
>Product Data Integration Technologies, Inc. 562-495-6509
>100 W. Broadway Suite 540 wburkett@pdit.com
>Long Beach, CA, 90802 USA http://www.pdit.com
>------------------------------------------------------------------
>
>
>
> > -----Original Message-----
> > From: Schoening, James R CECOM DCSC4I
> > [mailto:James.Schoening@mail1.monmouth.army.mil]
> > Sent: Tuesday, July 11, 2000 7:53 PM
> > To: 'standard-upper-ontology@ieee.org'
> > Subject: Call for vote on SUO Scope and Purpose
> >
> >
> >
> > SUO Participants,
> >
> > I would now like to call for a vote on the below Scope and Purpose.
> > The objectives of this vote are to:
> >
> > a. Cause a greater number of participants to review and understand
> > the Scope and Purpose.
> >
> > b. Solicit suggested changes and improvements.
> >
> > c. See if we have enough consensus and interest to proceed.
> >
> > To vote, send an email to me (not this list), by July 26, 2000, with
>
> > one of the following votes:
> >
> > YES (with or without comments)
> > NO (must have comments stating why you are voting NO)
> > ABSTAIN (with or without comments)
> >
> > IEEE does not define rules for voting within a Study Group, but
> > hopefully we'll be able to deal with most comments and reach maximum
> > consensus without strict voting rules.
> >
> > Please cast your vote.
> >
> > Jim Schoening
> > Chair, IEEE SUO Study Group
> > http://ltsc.ieee.org/suo
> >
> >
> > Scope of Proposed Project:
> > (The Scope describes what is being done, including the technical
>boundaries
> > of the project.)
> > This standard will specify the syntax and semantics of a general-purpose
> > upper level ontology. An ontology is a set of terms and formal
>definitions.
> > This will be limited to the upper level, which provides definition for
> > general-purpose terms and provides a structure for compliant lower level
> > domain ontologies. It is estimated to contain between 1000 and 2500 terms
> > plus roughly ten definitional statements for each term. It is intended to
> > provide the foundation for ontologies of much larger size and more
>specific
> > scope.
> >
> > Purpose of Proposed Project:
> > (The Purpose describes why the standard needs to be developed and who will
>
> > benefit.)
> > * The standard will be suitable for automated logical inference to
> > support knowledge-based reasoning applications.
> > * This standard will enable the development of a large (20,000+)
> > general-purpose standard ontology of common concepts to be developed,
>which
> > will provide the basis for middle-level domain ontologies and lower-level
> > application ontologies.
> > * The ontology will be suitable for "compilation" to more restricted
> > forms such as XML or database schema. This will enable database developers
>
> > to define new data elements in terms of a common ontology, and thereby
>gain
> > some degree of interoperability with other compliant systems.
> > * Owners of existing systems will be able to map existing data
> > elements just once to a common ontology, and thereby gain a degree of
> > interoperability with other representations that are
> > compliant with the SUO.
> >
> > * Domain-specific ontologies which are compliant with the SUO will be
> > able to interoperate (to some degree) by virtue of the shared common terms
>
> > and definitions.
> > * Applications of the ontology will include:
> > * E-commerce applications from different domains
> > which need to
> > interoperate at both the data and semantic levels.
> > * Educational applications in which students learn
> > concepts and
> > relationships directly from, or expressed in terms of, a
> > common ontology.
> > This will also enable a standard record of learning to be kept.
> > * Natural language understanding tasks in which a
> > knowledge based
> > reasoning system uses the ontology to disambiguate among likely
> > interpretations of natural language statements.
> >
> >
>
>
>-----------------
>Adam Pease
>Teknowledge
>(650) 424-0500 x571
-----------------
Adam Pease
Teknowledge
(650) 424-0500 x571