ISO 12620:2009

Terminology and other language and content resources - Specification of data categories and management of a Data Category Registry for language resources

ISO 12620:2009 provides guidelines concerning constraints related to the implementation of a Data Category Registry (DCR) applicable to all types of language resources, for example, terminological, lexicographical, corpus-based, machine translation, etc. It specifies mechanisms for creating, selecting and maintaining data categories, as well as an interchange format for representing them.

Terminologie et autres ressources langagières et ressources de contenu — Spécification de catégories de données et gestion d'un registre de catégories de données pour les ressources langagières

General Information

• Status: Withdrawn
• Publication Date: 09-Dec-2009
• Withdrawal Date: 09-Dec-2009
• ICS: 01.020 - Terminology (principles and coordination)
• ICS: 35.240.30 - IT applications in information, documentation and publishing
• Technical Committee: ISO/TC 37/SC 3 - Management of terminology resources
• Drafting Committee: ISO/TC 37/SC 3/WG 1 - Data categories
• Current Stage: 9599 - Withdrawal of International Standard
• Start Date: 19-Jan-2016
• Completion Date: 13-Dec-2025

Relations

• Revised: ISO 12620:2019 - Management of terminology resources - Data category specifications
• Effective Date: 28-Nov-2015
• Revises: ISO 12620:1999 - Computer applications in terminology - Data categories
• Effective Date: 15-Apr-2008

Standards Content (Sample)

INTERNATIONAL ISO
STANDARD 12620
Second edition
2009-12-15
Terminology and other language and
content resources — Specification of
data categories and management of a
Data Category Registry for language
resources
Terminologie et autres ressources langagières et ressources de
contenu — Spécification de catégories de données et gestion d'un
registre de catégories de données pour les ressources langagières

Reference number
©
ISO 2009
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.

©  ISO 2009
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail [e-mail]
Web www.iso.org
Published in Switzerland
ii © ISO 2009 – All rights reserved

Contents Page
Foreword .iv
Introduction.v
1 Scope.1
2 Normative references.1
3 Terms and definitions .1
3.1 Data elements and data categories .1
3.2 Data Category Registry.3
3.3 Data category specification components .4
3.4 DCR management.5
3.5 Roles.6
3.6 Data exchange .6
4 Role of data categories in language resource management .7
4.1 Overview.7
4.2 Variety of Data Category Selections (DCSs).8
5 Requirements for the implementation of a DCR for language resources .9
6 Registration Authority for the ISO/TC 37 DCR .10
7 Representation of data categories used in language resources .11
7.1 Introduction.11
7.2 Global Information class.11
7.3 Data Category classes .13
7.4 Administration Information Section .14
7.5 Documenting data categories .17
7.6 Conceptual Domain classes.20
7.7 Linguistic Section classes.21
7.8 Referencing a data category .23
7.9 Data Category Interchange Format .23
8 Management procedures for the ISO/TC 37 DCR.24
8.1 General organization.24
8.2 Roles and responsibilities.25
8.3 Thematic domain groups.25
8.4 Working procedure.26
8.5 Data Category Registry Board (DCRB) .28
Annex A (normative) Compact DC Reference RELAX NG Schema.30
Annex B (informative) Example of a DCIF Representation.31
Annex C (normative) Compact DCIF RELAX NG Schema .33
Annex D (informative) Alphabetical listing of definitions .38
Bibliography.40

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 12620 was prepared by Technical Committee ISO/TC 37, Terminology and other language and content
resources, Subcommittee SC 3, Systems to manage terminology, knowledge and content.
This second edition cancels and replaces the first edition (ISO 12620:1999), which has been technically
revised.
iv © ISO 2009 – All rights reserved

Introduction
Data associated with language resources are identified, collected, managed, and stored in a wide variety of
environments. Data items appearing in individual language resources are themselves referred to in this
International Standard as data categories, a designation commonly used in the environment of ISO Technical
Committee ISO/TC 37. Data categories as cited in ISO/TC 37 standards correspond to data element concepts
in the ISO/IEC 11179 series of standards, but differ slightly in terms of values defined. Differences in approach
among different kinds of language resources and individual system objectives inevitably lead to variations in
data category definitions and data category names. The use of uniform data category names and definitions
employed in resources within the same thematic domain (for example, among terminological resources,
lexicographic resources, annotated text corpora, etc.), at least at the interchange level, contributes to system
coherence and enhances the re-usability of data. Procedures for defining data categories in a given thematic
domain also need to be uniform in order to ensure interoperability among data categories, which becomes
problematic if they are defined in individual data category registries.
The creation of a single global Data Category Registry (DCR) for all types of language resources treated
within the ISO/TC 37 environment provides a unified view of the various applications of such a reference
resource. This universal registry is designed to facilitate a wide range of Data Category Selections (DCS)
needed in conjunction with all current or future standardization projects. ISO/TC 37 or any of its sub-
committees can resolve at any time to designate specific thematic domains to deal with the management of
those DCSs. The following thematic domains, among others, have been recognized as definable subsets of
the DCR for language resources:
⎯ “Terminology”: ISO 16642:2003 explicitly refers to a set of reference data categories for terminology
representation. Some of the data categories include general-purpose data management categories
1)
(for example, /source/ , /responsibility/, /date/, etc.) as well as linguistically oriented ones (for example,
/partOfSpeech/). Many of these data categories are relevant to a variety of different language resources,
not just to terminology management, and form the core of the DCR as described in this International
Standard;
⎯ “Semantic Content Representation” and “Lexical Resources”: the DCR serves as a reference for the
descriptors that are used throughout ISO/TC 37-related language resources, for instance, in terminology
management systems, at various levels of linguistic annotation (for example, morphosyntactic, syntactic,
and discourse levels), for lexical representation [natural language processing (NLP) lexicons, including
machine translation (MT) dictionaries, etc.], or for specific applications such as metadata for language
resources, query languages or multilingual data representation (for example, translation memories,
localization files, etc.);
⎯ “Language Codes”: ISO 639-1 and ISO 639-2 contain codes for approximately 650 languages. ISO 639-3
extends this number by an order of magnitude, with a clearer separation between the description of the
[1][2][3]
language and its coding proper . Including the reference set of language identifiers in the DCR in
response to the evolution of the ISO 639 family of standards provides an essential element of any
linguistic annotation or representation scheme.
⎯ “Lexicography”: the deployment of the DCR will include data categories for the description of
[4]
lexicographic data as cited in ISO 1951:2007 in order to ensure that the formats used for describing
lexicographical (SC 2), terminological (SC 3) and NLP-oriented (SC 4) data are comparable and
compatible.
1) Names that function as class names are capitalized. When a name functions as an attribute, it is represented in bold
face with the + convention: i.e. +administration record as opposed to Administration Record. This function is context-
dependent. Terminal values used with the data model appear in normal face bracketed by hyphens: -standardized name-.
Data category names are represented using the convention /source/. Data categories are themselves defined in the DCR,
not in this International Standard.
The DCR will eventually contain all ISO/TC 37 data categories, with their complete history, data category
descriptions, and attendant metadata. It is not, however, the intent of this International Standard to define an
ontology of language resources within ISO/TC 37. Nevertheless, the definition of the DCR has avoided any
choices that would hamper further work in this direction.
This document is intended to provide a background in the context of ISO/TC 37 on the various issues that
have to be considered in order to implement a global DCR that can be used for the full range of language
resources. More precisely, this document addresses the following issues:
⎯ the role of data categories for use with language resources;
⎯ requirements that have been identified with respect to information content and overall management;
⎯ a description of the organization of the DCR;
⎯ an interchange format for data categories, the DCIF (Data Category Interchange Format);
⎯ management procedures for the DCR.
Specific user-oriented instructions and procedures pertaining to the implementation and use of the DCR are
available on-line at http://www.isocat.org.

vi © ISO 2009 – All rights reserved

INTERNATIONAL STANDARD ISO 12620:2009(E)

Terminology and other language and content resources —
Specification of data categories and management of a Data
Category Registry for language resources
1 Scope
This International Standard provides guidelines concerning constraints related to the implementation of a Data
Category Registry (DCR) applicable to all types of language resources, for example, terminological,
lexicographical, corpus-based, machine translation, etc. It specifies mechanisms for creating, selecting and
maintaining data categories, as well as an interchange format for representing them.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 8601:2004, Data elements and interchange formats — Information interchange — Representation of
dates and times
ISO/IEC 11179-1:2004, Information technology — Metadata registries (MDR) — Part 1: Framework
ISO/IEC 11179-3, Information technology — Metadata registries (MDR) — Part 3: Registry metamodel and
basic attributes
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO/IEC 11179-1:2004 and the following
apply. Terms and definitions have evolved in the terminology community, represented here by citations from
ISO 1087-2, independently of the terminology of the metadata community, which results in slightly different
and at times overlapping concepts in the two communities of practice.
3.1 Data elements and data categories
3.1.1
data element
〈language resources〉 unit of data that, in a certain context, is considered indivisible
[ISO 1087-2:2000, 6.11]
NOTE In terminology work, an individual field, for example, /term/, in a single terminological entry has been viewed
as a data element and an instantiation of a data category (3.1.3).
3.1.2
data element
DE
〈metadata standards〉 unit of data for which the definition, identification, representation and value domain are
specified by means of a set of attributes
[ISO/IEC 11179-1:2004, 3.3.8]
3.1.3
data category
DC
result of the specification of a given data field
[ISO 1087-2:2000, 6.14]
EXAMPLE /partOfSpeech/, /grammaticalGender/, /grammaticalNumber/; the values associated with these items
(for example, /noun/, /verb/, /feminine/, /plural/, etc.) are also data categories according to this International Standard, but
values of this type are not viewed as data element concepts (3.1.4) in the ISO/IEC 11179 family of standards.
NOTE 1 A data category is an elementary descriptor in a linguistic structure or an annotation scheme (3.1.15).
NOTE 2 A data category corresponds closely, but not identically, to a data element concept in ISO/IEC 11179.
NOTE 3 In running text, such as in this International Standard, data category (3.1.3) names are set off using forward
slashes and italics. In some implementations, camel case is used instead of using white space between words in the data
category name.
3.1.4
data element concept
concept for which the definition, identification and conceptual domain (3.1.5) are specified independently of
any particular representation
[ISO/IEC 11179-1:2004, 3.3.9]
3.1.5
conceptual domain
set of valid value meanings
NOTE 1 Adapted from ISO/IEC 11179-1:2004, 3.3.6.
NOTE 2 The value meanings in a conceptual domain may be enumerated, further specified by additional constraints or
expressed via a description. For instance, the data category (3.1.3) /term/ is described by its definition and thus
constrained from properly containing, for example, contextual information or grammatical information, but it would be
impossible to enumerate all values associated with this data category.
3.1.6
value domain
set of permissible values
[ISO/IEC 11179-1:2004, 3.3.38]
3.1.7
complex data category
data category (3.1.3) that has a conceptual domain (3.1.5)
3.1.8
open data category
complex data category (3.1.7) whose conceptual domain (3.1.5) is not restricted to an enumerated set of
values
2 © ISO 2009 – All rights reserved

3.1.9
open conceptual domain
conceptual domain (3.1.5) associated with an open data category (3.1.8)
3.1.10
constrained data category
complex data category (3.1.7) whose conceptual domain (3.1.5) is non-enumerated, but is restricted to a
constraint specified in a schema-specific language or languages
3.1.11
constrained conceptual domain
conceptual domain (3.1.5) associated with a constrained data category (3.1.10)
3.1.12
simple data category
data category (3.1.3) with no conceptual domain (3.1.5)
3.1.13
closed data category
complex data category (3.1.7) whose conceptual domain (3.1.5) is restricted to a set of enumerated
simple data categories (3.1.12) making up its value domain (3.1.6)
3.1.14
closed conceptual domain
conceptual domain (3.1.5) associated with a closed data category (3.1.13)
3.1.15
annotation scheme
set of descriptors together with their syntax, semantics and condition of use, intended to provide descriptive or
interpretive information relevant to a language resource
[10]
NOTE TEI ODD (One Document Does it All) is an example of an annotation scheme.
3.2 Data Category Registry
3.2.1
Data Category Registry
DCR
set of data categories (3.1.3) to be used as a reference for the definition of linguistic annotation schemes
(3.1.15) or any other formats used in the area of language resources
3.2.2
data category specification
set of attributes used to fully describe a given data element concept (3.1.4)
NOTE The abbreviation “DCS” is associated with the Data Category Selection (3.2.3) and should not be confused
with the data category specification.
3.2.3
Data Category Selection
DCS
set of data categories (3.1.3) selected from the DCR (3.2.1)
NOTE 1 A DCS can represent the data categories used within a thematic domain (3.4.3) or a selection of data
categories used for a specific application or project. In the latter case, the DCS may draw data categories from more than
one thematic domain.
NOTE 2 A DCS can be expressed as a simple list of data categories, or it can be output in a form that contains the
entire content of their associated data category specifications (3.2.2), thus incorporating the full set of constraints
[11]
associated with the DCS. It can also be expressed using a schema notation such as W3C XML Schema or Relax
[12]
NG , which also comprises the list of data categories together with their associated constraints.
3.3 Data category specification components
3.3.1
DCR data model
logical representation of data structure and dependencies in the DCR (3.2.1)
[13]
NOTE 1 The DCR data model is represented as a UML class diagram .
NOTE 2 The definition above is based on ISO/IEC 11179-1:2004, 3.2.7, where “data model” is defined as a “graphical
and/or lexical representation of data, specifying their properties, structure and inter-relationships”.
3.3.2
Global Information
GI
technical and administration information applying to the entire data collection
[ISO 16642:2003, definition 3.7]
EXAMPLE Title of the data collection, revision history.
3.3.3
administration information section
class in a data category specification (3.2.2) pertaining to the submission, registration, balloting and
approval of data category specifications submitted to and maintained in the DCR (3.2.1)
3.3.4
registration group
class associated with the administration information section (3.3.3) that provides information related to the
Registration Authority (RA) (3.4.2) responsible for the administered item
3.3.5
submission group
class associated with the administration information section (3.3.3) that provides information on persons or
groups responsible for submitting the administered item
3.3.6
decision group
class associated with the administration information section (3.3.3) that provides information on the review
and balloting process associated with the administered item
3.3.7
stewardship group
class associated with the administration information section (3.3.3) that provides information on the
individual or group responsible for the maintenance of the administered item
3.3.8
description section
class pertaining to the data category (3.1.3) name and the data element concept (3.1.4) documented in a
data category specification (3.2.2)
NOTE Definitions, explanations, and notes comprise some of the kinds of information included in the description
class of a data category specification.
3.3.9
data element name
class in a data category specification (3.2.2) that lists and categorizes permissible names that can be
associated with the data category (3.1.3)
4 © ISO 2009 – All rights reserved

3.3.10
language section
class in a data category specification (3.2.2) that provides working language (3.3.12) equivalents for data
category (3.1.3) names and other descriptive information included in a data category specification (3.2.2)
3.3.11
linguistic section
class in a data category specification (3.2.2) that constrains the conceptual domain (3.1.5) for a given
object language (3.3.13)
3.3.12
working language
language used to describe objects
[ISO 16642:2003, 3.21]
3.3.13
object language
language being described
[ISO 16642:2003, 3.10]
3.3.14
name section
class in the language section (3.3.10) that lists variant names for the data category (3.1.3) treated in a data
category specification (3.2.2)
NOTE Variant names can be equivalents in other languages or names that may be used in the same language but in
related disciplines or working environments.
3.4 DCR management
3.4.1
Data Category Registry Board
DCR Board
DCRB
group of experts (3.5.3) designated by the participating (P) members of the Technical Committee whose duty
it is to ensure that the scope and the coherence of the Data Category Registry (3.2.1) are maintained
NOTE The DCRB has the status of a Validation Team (VT) according to Annex ST of the ISO Supplement to the
[9]
ISO/IEC Directives.
3.4.2
Registration Authority
RA
organization authorized to register data items and/or other information objects and to maintain them in a
repository
NOTE Typically these kinds of information objects can comprise codes, such as the language codes defined in the
ISO 639 family of standards, data categories, and other public identifiers. Registration Authorities are governed by
International Standards, but the repositories themselves are not generally included in published standards.
3.4.3
thematic domain
class of applications identified by the similarity of the data structures they need to manipulate
EXAMPLES Terminology, lexicography, morphosyntactic annotation.
3.4.4
thematic domain group
TDG
group of experts (3.5.3) in charge of selecting and maintaining the data categories (3.1.3) that are relevant
for a thematic domain (3.4.3)
NOTE Thematic domain groups have the status of the Maintenance Team (MT) according to Annex ST of the
[9]
ISO Supplement to the ISO/IEC Directives.
3.4.5
thematic domain profile
profile
representation within a data category specification (3.2.2) of the thematic domain (3.4.3) with which a data
category (3.1.3) is associated
NOTE A data category may have several thematic domain profiles, indicating that it is used by several thematic
domains. Until a data category specification is assigned to a TDG, the profile value is set to -private-.
3.5 Roles
3.5.1
chair of the Data Category Registry Board
chair of the DCR Board
chair of the DCRB
individual appointed by the ISO/TC 37 plenary who has the responsibility of administering the work of the
DCRB (3.4.1)
3.5.2
chair of a thematic domain group
TDG chair
individual appointed by the ISO/TC 37 sub-committee associated with a TDG (3.4.4) who has the
responsibility for administering the work of the TDG
3.5.3
expert
individual with special knowledge, skill, or other interest who registers to participate in the work of the DCR
(3.2.1)
3.5.4
judge
expert appointed by the chair of a thematic domain group (3.5.2) to participate in the approval process for
any given data category specification (3.2.2) or Data Category Selection (3.2.3) submitted for
standardization
3.6 Data exchange
3.6.1
Data Category Interchange Format
DCIF
export format for data categories (3.1.3) grouped as a Data Category Selection (3.2.3) designed to facilitate
their usability in external applications
3.6.2
snapshot
capture of the status of a data resource at a given moment in time
NOTE Data resources are frequently archived in the form of snapshots, which can then be identified as versions of
the resource.
6 © ISO 2009 – All rights reserved

3.6.3
persistent identifier
PID
unique Uniform Resource Identifier (URI) that ensures permanent access for a digital object by providing
access to it independently of its physical location or current ownership
4 Role of data categories in language resource management
4.1 Overview
Data category specifications identify the individual information units making up a data collection or annotation
scheme for a given language resource. A data category specification provides the formal representation of a
data category, which shall comprise the specific attributes that document it (for example, the data category
name, definition, examples, comments, etc.). It shall also provide the context for its creation and management
within the DCR. Groups of data categories subsetted from the global set making up the DCR comprise Data
Category Selections (DCS). As specified in ISO 16642, Terminological Markup Framework (TMF), a DCS
shall define, in combination with a data model, the various constraints that apply to given thematic-domain or
application-specific information structures or interchange formats.
Figure 1 shows possible uses for a DCS. Depending on the application involved, such a DCS may be merely
a list of data categories that points back to the complete specifications in the DCR, or it can be represented by
a complete subset or even superset of the DCR, comprised of such a list plus the definitions and constraints
associated with the individual data category specifications.
From a wider perspective, a formal model for representing data categories shall account for the fact that apart
from pure computer use, a data category specification can be intended for human use as well. For instance,
such specifications can form the core of a DCS, which can be published either as a paper document, made
available as an electronic resource, or identified as a subset of the ISO/TC 37 DCR. Typically, the designers
of a given markup language or data management system will query the DCR in order to create their individual
application-specific DCSs by selecting a subset of data categories from the global DCR. Finally, providing a
precise description of the data categories used within a given data collection in reference to the DCR allows
for a quick diagnosis of the compatibility of this collection with any other particular computer application and
thus can act as metadata for this collection.
Figure 1 also presents the notion of a DCS, for example, the choice of a specific set of data categories taken
from the global DCR for use in a given thematic domain within the framework of language resources, or in a
specific application. The diagram exemplifies the various roles of a DCS in the process of defining and using
any linguistic annotation scheme. Viewed from this perspective, a DCS is primarily intended to contribute to
the specification of the DCR annotation scheme in combination with the data model that expresses the
general organization of the DCR. This kind of selection guarantees a certain degree of interoperability
between two or more data structures by facilitating the comparison of the selected data categories as well as
the constraints imposed on them, for instance the nodes in the DCR data model that correspond to the
positions where each category is allowed to occur in the individual data structures, such as in specific
applications or annotation schemes. In this scenario, the DCS for each of the structures in question can be
[12] [11]
expressed as, for example, a Relax NG or W3C XML Schema , and XSL filters can be used to output
relevant data from the Data Category Interchange Format (DCIF) in alternative formats (see 7.9).
In addition, the DCS can be seen as a documentary source for the linguistic annotation scheme in question.
Indeed, because it contains the list of all data categories that can be used in conjunction with the annotation
scheme, it is probably the best source of information for potential users or implementers who want to know
whether a given data category corresponds to their needs.
Furthermore, the DCS can be attached (or referenced) in any data transmission process to provide the
receiver with all the information needed to interpret the content of the information being transmitted. In
particular, this procedure should allow linguistic data expressed in various kinds of XML representations to be
sent or received in the most transparent way.
Figure 1 — Role of Data Category Selections
in the context of the definition of linguistic annotation schemes

4.2 Variety of Data Category Selections (DCSs)
Figure 2 illustrates the relationship between individual data category specifications, the DCR, and any one of
the possible DCSs that represents a subset of the DCR. The largest circle represents the entire collection of
data category specifications included in the DCR, and the smaller internal circles represent DCSs which are
subsets of the DCR. The very small, individual background patterns in the drawing correspond to individual
data category specifications, each describing a given data category concept using the data category attributes
set down for the DCR with reference to the attributes defined in ISO/IEC 11179. For instance, one small
pattern element might represent the data category specification for the /term/ data category.
Some data categories included in the DCR are pertinent to a single thematic domain within the broader field of
terminology and other language and content resources. For instance, a /conceptIdentifier/ is probably unique
to terminological resources (although not prescriptively), or a /senseNumber/ is probably specific to
lexicographical resources. On the other hand, many data categories, frequently those of a strictly linguistic
nature such as /partOfSpeech/, /grammaticalGender/, /grammaticalNumber/, etc., are common to a wide
variety of resources. To be sure, these categories may not always have the same function in different thematic
domains, but they nevertheless represent the same essential concept in different kinds of resources. Hence
each thematic domain shall contribute all its data categories in the form of data category specifications to the
global DCR, while at the same time identifying those data categories that it shares with other kinds of
resources. The subset of data categories and their specifications used in a thematic domain shall comprise a
domain-specific DCS taken from the DCR.
As noted, the oval shapes in the diagram represent such DCS subsets. A further subset can be selected from
one or more thematic domain DCSs for use in a given application or collaborative environment. The octagon
shown in Figure 2 represents an example of an application-specific subset. This particular application is wholly
contained within the DCS for terminological data categories, so it is apparently designed for use with a
8 © ISO 2009 – All rights reserved

terminological application, although some of the data category specifications contained in this subset are
common to several different kinds of language resources. It should thus be noted in this context that some
applications can indeed draw data categories from several thematic domain DCSs. Theoretically, it is also
possible for a DCS to be in part a superset if it happens to include categories that are not currently part of the
DCR. In such cases, it is recommended that designers or users of such a set register their new data
categories in the DCR.
Figure 2 — Individual application-oriented DCS viewed in the context of various TDG-oriented DCSs
5 Requirements for the implementation of a DCR for language resources
This clause outlines the basic requirements that a DCR should fulfil in order to support standardization
activities within ISO/TC 37.
The ISO/TC 37 DCR shall:
⎯ Provide a reference repository for data categories and related information for all the existing or future
standards in ISO/TC 37 that involve data modelling or data interchange.
⎯ Be available online, free of charge.
⎯ Register existing practices by associating each data category with the way it is implemented in specific
projects or initiatives. This may consist in registering various types of encodings, from basic codes
[15]
(for example, ‘f’ for feminine in EAGLES morphosyntactic descriptions) to actual XML representations.
⎯ Provide names and definitions in a variety of languages.
⎯ Describe the use of each data category in a variety of object and working languages. This may consist of
a specific definition (for instance when the data category has a slightly different application scope), some
usage notes, examples, or list of values (for example, the value domain of /gender/ is {/masculine/,
/feminine/} in French, and {/masculine/, /feminine/, /neuter/} in German).
⎯ Describe the use of data categories in a variety of resources, where necessary.
⎯ Associate administration information with each data category so that it is possible to trace the submission,
acceptance or revision of the data category.
⎯ Associate each data category with one or several profile values corresponding to the thematic domains
where the data category is relevant (for instance, /partOfSpeech/ is relevant for both terminology and
lexicography).
⎯ Provide a mechanism by which a thematic domain group in ISO/TC 37 can submit a group of data
categories relevant to its scope of activities.
⎯ Provide private workspaces where individuals and working groups outside ISO/TC 37 can generate or
upload their own data category specifications and selections, to publish and share them, and, when
desired, to submit data categories for registration and approval.
⎯ Be updated on a regular basis by integrating, according to defined rules, proposals from experts in the
field.
⎯ Be implemented in accordance with the main principles stated in the ISO/IEC 11179 family of standards.
⎯ Provide stable worldwide accessibility by maintaining distributed mirror sites.
⎯ Provide for persistent identifiers by introducing a system of stable references for individual data
specifications.
⎯ Maintain secure, best practices for archiving procedures.
⎯ Provide ISO with periodic data snapshots of those data categories that have been approved following
procedures spelled out in Annex ST of the ISO Supplement to the ISO/IEC Directives and Clause 8 of this
International Standard, with the understanding that ISO shall make this subset of the DCR available
through its standard practices without compromising the availability of these items within the environment
of the DCR itself.
6 Registration Authority for the ISO/TC 37 DCR
The ISO/TC 37 DCR shall be implemented under the auspices of a Registration Authority (RA), as defined in
ISO/IEC Directives, which has the duty to register data categories in accordance with the rules defined in this
International Standard.
2)
The Technical Management Board has appointed the Max Planck Institute for Psycholinguistics to be the
Registration Authority for the ISO/TC 37 DCR.
The RA shall maintain the DCR as a web service at http://www.isocat.org. All user functions and support shall
be available through this website.

2) At the time of publication of this International Standard, the ISO/TC 37 DCR Registration Authority is held by:
Max Planck Institute for Psycholinguistics, Nijmegen, The Netherlands.
ISO maintains an online listing of maintenance and registration authorities at http://www.iso.org/rama. Users are
encouraged to consult this listing for the most up-to-date information concerning maintenance agencies and registration
authorities.
10 © ISO 2009 – All rights reserved

7 Representation of data categories used in language resources
7.1 Introduction
This section describes the data model of the DCR for ISO/TC 37. The data model is specified in Unified
[13]
Modelling Language (UML) , extended where necessary with additional constraints expressed in the Object
[16]
Constraint Language (OCL) . The complete data model is shown in an overview in Figure 3, broken out in
three parts in Figures 4, 5 and 6, where the Data Category serves as the linking class for all figures.
The DCR shall consist of two main classes:
⎯ Global Information (GI);
⎯ one or more Data Category (DC) specifications.
Each data category specification shall consist of two mandatory classes:
⎯ one dedicated to the administration and identification of the data category (the Administration Information
Section);
⎯ one dedicated to the documentation of the data category in one or more working languages and zero or
more names as used in a given database, format or application (the Description Section).
Further enhancements may be made to the data category specification depending on the exact data category
type. These include the following classes:
⎯ one or more classes which describe the conceptual domain of the data category (the Conceptual Domain
and subclasses);
⎯ one or more classes which describe the conceptual domain and/or the use of the data category in the
context of a specific object language (the Linguistic Section and subclasses).
Further classes associated with these main classes are described in more detail in the following subclauses.
7.2 Global Information class
The DCR shall make available the following items of information reflecting Global Information:
⎯ the name and complete contact information of the Registration Authority (complete address,
telephone and fax numbers, e-mail contact address);
⎯ the date when the RA was appointed and other critical dates concerning versions of the RA;
⎯ the name of the person primarily responsible for administration of the RA;
⎯ historical information concerning previous RAs in the event that responsibility for the RA is
transferred to another entity;
⎯ short forms or abbreviations used in the DCR to reference the current RA and any previous RA(s)
that may have been responsible for the registry;
⎯ names and affiliations of members of the DCRB;
⎯ information on the RA agreement between ISO and the RA;
⎯ the DCR mission statement;
⎯ a statement of legal responsibilities and restrictions.
This International Standard does not impose any additional specific constraints on this class. It is the
responsibility of the RA and the DCRB to provide technical information as deemed fit.
U inheritance relationship from subclass to superclass
 composition relationship between two classes
š aggregation relationship between two classes
Figure 3 — Overview of the DCR data model (see Figures 4, 5, and 6 for details)

12 © ISO 2009 – All rights reserved

U inheritance relationship from subclass to superclass
 composition relationship between two classes
š aggregation relationship between two classes
Multiplicity values: 0.1 = 0 or 1; 0.* = 0 or more; 1 = 1 and only 1; 1.* = 1 or more instances may occur.
Figure 4 — Administration part of the DCR data model
7.3 Data Category classes
7.3.1 Data Category subtypes
Two primary subtypes of Data Category are maintained in the DCR and distinguished as follows:
a) Complex Data Categories, such as /term/ or /grammaticalGender/;
b) Simple Data Categories, such as /masculine/, /feminine/, etc.
These two basic types of data category are represented in the DCR model as two subclasses of Data
Category. They are shown in Figure 6 in conjunction with the Linguistic Section (see 7.7).
Data Category contains the following attribute:
⎯ +pid [1]: the persistent identifier (PID) of the data category; for standardized data categories, the RA shall
...