SUO: RE: SUO-STAT Program wants to work with SUO documents
Dear Jim,
Here is my first attempt at answering the BIKE questionnaire as is.
Strawman Evaluation Questions for Common Upper Ontologies
Ontology Name: ISO 15926 - Integration of Lifecycle Data
Brief Description:
ISO 15926 has been developed in ISO TC184/SC4 - Industrial Data,
by the Oil and Gas and Process Industries through a consortium,
EPISTLE, which itself has consortia as members. More than 100
companies have been involved in its development, which began in 1993.
ISO 15926 is an implementation of the integration architecture described
in ISO TS 18876-1:2003 - Integration of industrial data for exchange
access and sharing. There are a number of elements, in various stages
of formal standardisation:
- Data Model, this contains the top 200 concepts of the ontology, which
includes a meta model for extending the ontology through what is known
as a Reference Data Library. The data model is written using the
EXPRESS language specified in ISO 10303-11:1995.
- Reference Data Library, this currently contains some 20,000 concepts
from the engineering domain, specifications are being developed for the
maintenance of the Reference Data Library as an ISO Register, allowing
timely extension and access over time.
- XML Exchange Templates and Facade, this supports the exchange of
specific messages for specific purposes in terms of the ISO 15926 Data
Model and Reference Data Library, enabling concise yet specific exchanges,
and hiding detail. The Facade provides an internet based approach for the
controlled integration and sharing of data across a number of
collaborating partners.
1. Maturity: (How ready is it to use now? What capabilities have already
been demonstrated? Time and resources needed to start using? Potential
for improvement.)
The ontology has been developed over a 12 year period, and has already
been used to support multi billion dollar capital projects in the Oil
and Gas industry, integrating and exchanging engineering design data
between contractors and owner operators. It has also found use in the
Rail Industry.
There is commercial software available that can embed the ontology
directly for integration purposes.
The integration architecture definition has been standardised through
ISO as a Technical Specification:
- ISO TS 18876-1:2003 - Overview and Architecture
- ISO TS 18876-2:2003 - Methodology
The data model has been standardised as ISO 15926-2:2003.
The Reference Data Library has been developed and been in use for some
years, and the standardisation process is due to start shortly as an
ISO Technical Specification (ISO 15926 parts 4, 5, and 6).
The XML Templates and Facade have been implemented, and standardisation
is due to start shortly as a Technical Specification, ISO 15926-7.
Historically pilot projects on a small but real problem have cost about
US$1-2m. On a pilot project I was involved in in 1997 first benefits were
delivered within 3 months of an 18 month development.
The architecture is designed to support improvement in both scope and
depth.
2. Robustness: (Heavy weight vs. light weight ontology features?
Potential for improving robustness? How well will it handle known
requirements, such as those listed in SUO Scope and Purpose.)
The main distinguishing feature of ISO 15926 is that it is founded on
a specific metaphysical view of the world, and this is applied rigorously.
The metaphysical view applied is known as four dimensionalism which sees
individual objects as extended in space as well as time. This builds in
the approach to managing change, something that can cause problems. This
approach has been chosen specifically because of the lack of exceptions
that arise, and the consistency it gives in representation. It is possible
to translate into other world views. Other foundation concepts are mainly
from mathematics.
Whilst considerable care has been taken in identifying and defining
foundation concepts, there is still work ongoing to extend the
axiomatisation of them.
3. Potential for broad acceptance: (How well will it support maximum
number of domains?)
Whilst the ontology has a particular world view point, one reason for
selecting this was its neutrality with respect to domain. The whole
architecture is designed around bringing information together from
different domains.
4. Language Flexibility: (What ontology language is it in? How stable is
language?If desired, could it be written in a different ontology
language?)
The data model in ISO 15926-2 is defined in EXPRESS (ISO 10303-11). The
data model has already been translated into XML-Schema for use in the
XML Templates, and there is current work ongoing to translate it into OWL.
A meta-model of Common Logic has been developed in EXPRESS, and integrating
this with the current ontology should enable the support of Common Logic
compliant languages.
5. Ownership/Cost/Changes: (Who owns it? Any proprietary restrictions on
use? Any charges for utilization? How will it get changed and who
controls the changes? Is it being developed by a Standards Developing
Organization?
Most of the work described above is ISO Copyright and is subject to their
usual terms and conditions. Most development work is in the public domain,
but some extensions to the Reference Data Library have been made that are
private, either to particular companies, or particular consortia. For
example POSC Caesar has developed some 40,000 concepts in addition to the
20,000 in the public Reference Data Library that are currently private
(though I understand there are plans to make these public).
Clearly there are costs to publishing and maintaining an ontology. This has
been much discussed in EPISTLE. The consensus seems to be that for concepts
that are properly public, that content should be free at the point/time of
use and hence costs born by those who see benefit in the concepts being
publicly available, i.e. by ontology development projects that wish to add
their concepts to the Register, and have a business justification for this.
Integration of new concepts is a skilled task, and this needs to be a
controlled process independent of ontology development projects, and hence
provided as a service to projects from a trained and qualified group of
people. For particular domains the concept of "peer groups" for reviewing
proposals has been suggested.
Work is also going on in the SUO WG, particularly looking at increasing the
level of axiomatisation of the ontology.
6. Domain Friendly (How easy to develop domain ontologies based on upper
ontology?)
The integration architecture specifies how you add concepts from an
existing domain ontology into the ISO 15926 ontology, templates can be
developed to support the integration of data in the domain ontology with
the integraion ontology, and hence any other ontology integrated to it.
Equally, if a domain ontology is required but does not yet exist, it can
be created by adding the (missing) concepts directly into the integration
model (Reference Data Library).
Outline methodologies for this are provided in ISO TS 18876-2.
Matthew West
Principal Consultant
Shell Information Technology International Limited
Shell Centre, London SE1 7NA, United Kingdom
Tel: +44 20 7934 4490 Mobile: +44 7796 336538
Email: matthew.west@shell.com
Internet: http://www.shell.com
http://www.matthew-west.org.uk