Quick Scroll To:
CONCEPTS
Every concept has a definition, two ways to denote that concept (a
label, a code), and ties to concepts in other documents (standards, specifications,
dictionaries, publications).
- The definition field of an EDCS Dictionary Entry is a
precise statement of the nature, properties, scope or essential qualities
of a concept embodied in the entry. Definitions are complete and concise,
and may include the name or names for the concept as part of the definition.
- The label field of an EDCS Dictionary Entry is a compact
and human-readable designator that is used to denote a concept. Labels
may include the name or names for the concept. Labels:
- uniquely denote a concept within a dictionary,
- are a succinct expression of the concept it denotes,
- are represented as a character string, and
- are human readable.
- The code field of an EDCS Dictionary Entry is a compact,
and not necessarily human-readable, designator that is used to denote a
concept. Codes:
- uniquely denote a concept within a dictionary,
- are represented as an integer, and
- are assigned sequentially in increasing order within an EDCS Dictionary,
beginning at 1.
- There is a one-to-one relationship between labels and codes in the
same EDCS Dictionary. Therefore, a label and a code may be used interchangeably
to denote the same concept.
- Concepts in two documents (the EDCS and some external document) may
be related in several different ways. Reference types, and
references, establish the relationship between an EDCS concept
and an external concept.
CONCEPT TIES
To improve unambiguous communication across disparate domains, the
goal is to bring together (consistently) concepts that are seen as "authoritative"
in those domains. Typing references helps identify the degree to which
an EDCS concept is viewed as "authoritative" in another community.
It also may serve as a starting point for developing mappings from the
specialized languages of communities to the cross-community language of
the EDCS. Two types of reference (normative and informative) are insufficient
to capture all of the possible relationships between the EDCS and another
document. In the following, a specification is a document that meets
the requirements to be normatively referenced by the EDCS, and the owner
of such a specification is the organization responsible for maintaining
the specification.
- Prescriptive reference (PR): The concept in the EDCS
is the same concept as defined in an external specification. The definition
of the concept in EDCS may be different from the definition in the specification
because of modifications to:
- include missing (implied) context not present in the definition in
the specification; or
- to match the style and structure of other definitions in the EDCS.
- In this case, a citation is provided in the references field.
- Authoritative reference (AR): The concept in the EDCS
prescriptively references a concept from an external specification, and
an agreement has been established between the owner of the EDCS and the
owner of that specification to ensure that the concept authoritatively
referenced is not changed without the agreement of the owner of the EDCS.
In this case, a citation is provided in the references field.
- Dictionary reference (DR): The concept defined by a single
word or a short phrase in the EDCS is the same concept as defined by a
specified entry in the New Shorter Oxford English Dictionary (NSOED). The
NSOED (and the location within it) is appropriately cited in the references
field.
- Informative reference (IR): The concept from the EDCS
is related to, or derived from, concepts in one or more external documents.
In this case, one or more citations are provided in the references
field.
- Non-referenced (NR): No satisfactory prescriptive, authoritative,
dictionary or informative reference is available in another document for
re-use as a concept in the EDCS. The meaning of the definition of the concept
is determined by using the appropriate dictionary definitions of the words
comprising the definition, considering the context in which each word is
used. In this case, no citation is provided in the references field.
CONCEPT REUSE
Concepts may be used within the definition of other concepts. This
creates a web of explicitly inter-related concepts, reduces the complexity
of individual definitions, and ensures that concepts in, and across, dictionaries
"fit together".
- If an EC is used in the definition of another concept, this use is
represented in the definition by placing the label for that EC in single
angle brackets (<�>) in the definition.
- If an EA is used in the definition of another concept, this use is
represented in the definition by placing the label for that EA in double
angle brackets (<<�>>) in the definition.
CLASSIFICATIONS and ATTRIBUTES
Classifications are typings-of-objects (in the EC Dictionary);
attributes are states-of-objects (in the EA Dictionary). Attributes
are, in effect, measures. Measures are typed in order to define the value
storage mechanism being used. Attributes have one of the following value-types
assigned:
- real: A real number or a real interval.
- integer: An integer or an integer interval.
- count: A cardinal number or a cardinal number interval.
- index: An integer used for identification.
- string: A character string.
- constrained string: A string whose format and
values are constrained in the definition of the EA.
- key: A string used for identification.
- enumeration: One of a finite set of mutually exclusive
values.
- logical: An enumeration, representing the true
and false values in a two-valued logic system, with values TRUE (1) and
FALSE (2).
- null: An enumeration with one value, NULL (1),
representing no information.
Real-, integer-, and count-typed attributes may
be either a single value or an interval (value-pair). Real-typed attributes
have units of measure, which are captured in the EU Dictionary.
.
Measures may be missing, withheld, many-valued (for a partitioned range),
etc. These metadata are captured using concepts in the EM Dictionary.
ATTRIBUTE ENUMERANTS
While some measures are continuously valued (real, integer,
count), others are discretely valued (e.g., null, logical,
enumeration). While the discrete values of null and logical
can be pre-defined, the discrete values of enumeration depend on
the attribute concept -- hence attributes of value-type enumeration
must declare their value domain (e.g., attribute colour has the
value domain of {red, orange, yellow, green,
...} ). Concepts in the value domain of a measure (attribute) have their
own EE Dictionary, and all of the mechanisms applied to classifications
(which are typings-of-objects, as opposed to typings-of-discrete-values-of-measures)
are applied to attribute enumerants.
UNITS OF MEASURE and VALUE CONVERSION
Units characterize the measurement of quantities. Some quantities may
have more than one appropriate unit-of-measure. Unit equivalence classes
(in the EQ Dictionary) identify units that are comparable, and therefore
values characterized by those different units of measure can be interconverted
(e.g., foot-to/from-metre). A functional interface is defined, and
implemented, to support such value conversions. While most units are compliant
with the SI (Système International), some predate the SI. While
they are supported, they are also deprecated (e.g., statute mile,
hectare, knot, rad). Attribute values may be scaled in order to better
support a community of interest (e.g., kilometres, hectopascal,
milligrams). The ES Dictionary includes such concepts.
ORGANIZING CONCEPTS
Concepts can be organized into groups (in the EG Dictionary)
sharing a common "scope of application" or "theme"
(e.g., living organisms, vehicles, space, land transportation).
Concepts can be relevant to multiple groups (e.g., the U.S. Space
Shuttle is relevant to both the vehicles and space groups). Classifications
and attributes can be (multiply) grouped. Groups aid in searching the EDCS
to find concepts of relevance to a coding task. Membership in a group does
not affect/change the concept.
Groups are members of organizational schema (in the EO Dictionary),
such that every classification or attribute concept is required to be a
member of at least one group in that schema. This ensures that there are
no "holes" in the coverage of the groups comprising a schema.
Many schema (and their groups) can be defined; each from the perspective
of a community-of-interest. Only one such schema is currently defined,
general, however more can be easily added.
REGISTRATION and DEPRECATION
In order to support continuous evolution and growth of the EDCS, a
low-overhead process of registration has been defined, allowing
for the timely addition of new concepts to the EDCS. Additionally, a slower,
measured process of deprecation has been defined for removing out-dated
or inappropriate concepts from the EDCS.
PROFILES
In order to support the specification of rational subsets of
the EDCS for use within a community, and constraints on the use
of specific combinations of concepts from the EDCS, a profile mechanism
has been defined. The base profile includes all concepts from all
EDCS Dictionaries, and includes no constraints. Additional profiles may
be defined through the registration process.
Return to: Top of this
Page
|