27 No. 3
An Ontology1 on Property for Physical, Chemical, and Biological Systems
by René Dybkaer
APMIS Suppl. 2004 (117):1-210. PMID: 15588029
ISBN 87 9900100
Recently, one of the most distinguished scientists in laboratory medicine, Dr. René Dybkaer, defended a dissertation at the University of Copenhagen. According to Dr. Anders Kallner, former president of the Chemistry and Human Health Division, this remarkable work will have implications far beyond clinical chemistry. Dybkaer, who turned 80 recently and is still very active, was once chairman of the IUPAC Commission of Clinical Chemistry.
It is typical for scientific disciplines, such as laboratory medicine, to need to present the outcomes of examinations of a multitude of properties of many different types of objects. The special language used in each discipline is mostly found in texts on metrology, notably the International Vocabulary of Basic and General Terms in Metrology (VIM 1993). These sources, however, are partially conflicting, have not been established in a coherent fashion using adequate terminological procedures, and by definition relate only to quantities, ignoring all properties without magnitude.
Drawing on a variety of texts, it proves possible to form a comprehensive domain ontology around “property” with systematic definitions and terms constructed by using the tools provided in recent ISO International Standards on terminology work, supplemented by a few extra concepts to clarify the discussion. The backbone of the concept system is furnished by “system” including “component” and associated with “property”, further connected by consecutive associative relations to “examination procedure”, “examination”, “property value”, and “property value scale”. The description is supplemented by “kind-of-property” to indicate a generic concept for mutually comparable properties.
From each of these vertebral superordinate concepts, specific concepts are intensively defined and given systematic terms using the modifiers ‘nominal’, ‘ordinal’, ‘differential’, and ‘rational’. Each of the ensuing coordinate concepts corresponds to a typical allowed mathematical and statistical treatment of the property values (i.e., fundamentally the respective applicability of the operators (=, #), (>, =, <), (+, -), and (x, :) in four increasing sets obtained by cumulation to the right.
For the narrower field of metrology, ignoring nominal properties, an alternative concept system is presented based on “quantity”, which covers all properties having magnitude. The following subordinate level separates “ordinal quantity” and “unitary quantity”, the latter covering in its turn the specific concepts “differential quantity” and “rational quantity”.
Supplementary concepts comprise “examinand” and “measurand”, “examination method” and “measurement method”, “examination principle” and “measurement principle”, “true property value” and “examined property value”, as well as the well known “measurement”, “quantity value”, and “numerical value”. Also “examination result” and “measurement result” with respective “examination uncertainty” and “measurement uncertainty”, as well as “quantity value scale” are defined.
For differential and rational properties, the concept “metrological unit” is essential to their measurement and expression of unitary quantity values, although the term for a unit does not indicate the kind-of-quantity. Furthermore, great advantage accrues from creating a “system of metrological units”, preferably a universal “coherent system of metrological units”. Currently, this is the “International System of Units,” SI, formed from specific concepts of “base metrological unit” (seven in all) and “coherent derived metrological unit” (a large number), with multiples and submultiples obtained by SI prefixes; these concepts are defined as well as “off-system metrological unit” and “in-system metrological unit”. The lack of a special term for “metrological unit one” is discussed with support for the proposal ‘uno’ by the CIPM Consultative Committee for Units. The BIPM Brochure on the SI has no formal definitions, but a concept system can be inferred from the text. Its insistence that “SI unit” only comprises base and coherent derived metrological units creates practical problems; a modified concept diagram that solves them is suggested.
The further abstraction of “unitary kind-of-quantity” via “metrological unit” leads to “metrological dimension”, which is a powerful tool in “dimensional analysis” based upon the defining algebraic relationship between unitary kinds-of-quantity. As is the case for metrological unit, a metrological dimension does not identify a unique corresponding unitary kind-of-quantity. Also “base metrological dimension” and “derived metrological dimension” can be defined, as well as the much discussed “derived metrological dimension one”.
The designations for singular properties have been given a syntactic structure and semantic rules during 40 years of work within IUPAC, the IFCC (International Federation of Clinical Chemistry and Laboratory Medicine), and lately the CEN (Comité Européen de Normalisation). It is proposed that the tripartite general concept comprising “system”, “component”, and “kind-of-property” enters into the definition of “dedicated kind-of-property”. It is shown that the CEN Technical Committee 251 semantic model for definitions and terms of specific concepts utilizing generative patterns can be applied in forming representations of such dedicated kinds-of-property.
Finally, it proves possible to regard representations of individual concepts under “property” and “property value” in a mathematical and logical formalism operating according to the definition of a given dedicated kind-of-property. Examples are given for relation or function within Set Theory or Object-Oriented Analysis. The Set-Theoretical relation appears most flexible and leads to a definition of “Set-Theoretical representation of dedicated kind-of-property”.
The outcome of the investigation is a concept system with definitions and systematic terms permitting unambiguous description of dedicated kinds-of-property (except those involving vectors and tensors), designations for singular properties, and examination results encountered in laboratory work.
1 Ontology, n. the branch of metaphysics that studies the nature of existence (Random House Dictionary)
Ontology n. philosophy concerned with the nature of being (Oxford Dictionary)
Dybkaer is working at the Department of Standardization
in Laboratory Medicine at the Frederiksberg Hospital and at
the Faculty of Health Sciences at the University of Copenhagen.
Dybkaer has been an active member in IUPAC since the 1970s,
and is currently involved with the Interdivisional Working
Party for Harmonization of Quality Assurance (of the Analytical
Chemistry Division) and the Subcommittee on Nomenclature,
Properties, and Units in Laboratory Medicine (of the Chemistry
and Human Health Division).
last modified 21 April 2005.
Copyright © 2003-2005 International Union of Pure and
Questions regarding the website, please contact [email protected]