9 th open forum on metadata registries harmonization of terminology, ontology and metadata 20th –...

9th Open Forum on Metadata RegistriesHarmonization of Terminology, Ontology

and Metadata20th – 22nd March, 2006 , Kobe Japan.

ISO TC 37 Tutorial

Gerhard Budin

University of Vienna

Chair, ISO TC 37/SC 2

<[email protected]>

9th Open Forum for Metadata Registry, Kobe, 2006

Overview

ISO TC 37 – foundations and current work Terminological semantics for semantic

interoperability R&D for creating new standards – LIRICS Application projects + pre-normative R&D –

ADNOM (Meta-data) Modeling approaches in TC 37 –

LMF Conclusions


TC 37 Matrix

SCs SC 1 SC 2 SC 3 SC 4Topics

Cultural/ling. diversity 22134 1261612199

Metadata/coding 1087 639 12620 2182912615 1087-2

Semantic interoperability 860 1951 12200 24613704 639 16642

12616Lang. res. management 860 1951 12618 24613Terminology work 704 22128Terminography 15188Lexicography 10241

1219912616


The TC 37 „meta-language“

ISO 1087 contains the axiomatic – philosophical and linguistic – foundations of terminology work based on (traditional ) terminology theory

ISO 704 operationalizes these axioms into rules and principles and provides a basic methodology

ISO 12620 specifies how to use and register data categories

ISO 639 specifies how to use and register language codes


LIRICSLinguistic Infrastructure for Interoperable Resources and

SystemsGOALS: LIRICS provides a common standards framework for

language engineering by translating requirements from language industry into ISO standards on the basis of ongoing R&D work

LIRICS provides input, on the basis of the cooperation and interaction between research consortia and industry groups, to ongoing standards work in ISO/TC 37, mainly focusing on lexicons, morpho-syntax, syntax and semantic content. These standards will be accompanied by a set of test suites in nine European languages to facilitate their implementation and an open source implementation platform allowing common-format, multi-lingual language processing compatible with legacy systems and tools


Primary resourcesTexts, spoken data,

multimedia information[TEI, MPEG7, TMX,

XHTML, etc.]

NLP structuresLinguistic annotations

TokenisationMorpho-Syntactic Tagging

Chunks (e.g. Named Entities, etc.)Deep syntactic structures

Co-references etc.[Eagles, ISLE,Multext/Multext-East

CES, MATE, Whiteboard]

Knowledge structuresHierarchies of types

Relations between concepts[Topic Maps,

RDF/RDFS/OWL]

Lexical structuresTerminologies

Morphological lexicaSyntactic lexicaTransfer lexica

[ISO 16642, TBX, OLIF, Genelex/Simple/ISLE]

Meta-data[Dublin core, TEI,

OLAC, IMDI, MPEG7]

Access protocols[Corba, SOAP]

LIRICS domain of impact

LIRICS scope

Primary resourcesTexts, spoken data,

multimedia information[TEI, MPEG7, TMX,

XHTML, etc.]

NLP structuresLinguistic annotations

TokenisationMorpho-Syntactic Tagging

Chunks (e.g. Named Entities, etc.)Deep syntactic structures

Co-references etc.[Eagles, ISLE,Multext/Multext-East

CES, MATE, Whiteboard]

Knowledge structuresHierarchies of types

Relations between concepts[Topic Maps,

RDF/RDFS/OWL]

Lexical structuresTerminologies

Morphological lexicaSyntactic lexicaTransfer lexica

[ISO 16642, TBX, OLIF, Genelex/Simple/ISLE]

Meta-data[Dublin core, TEI,

OLAC, IMDI, MPEG7]

Access protocols[Corba, SOAP]

LIRICS domain of impact

LIRICS scope

SKOS


LIRICS Consortium INRIA, Nancy (FR) DFKI, Saarbrücken, (DE) UFSD, Univ. of Sheffield (UK) CNR–ILC, Pisa (IT) Univ. Vienna (AT) UTiL, Tilburg Univ. (NL) MPI, Max Planck-Inst. (DE) Univ. of Surrey (UK) IULA- UPF Univ. Pompeu Fabra, Barcelona (ES)

Industrial Advisory Group: Morphologic, Systran, Polderland, Trados, Acrolinx, BT Exact, EADS CCR, Esteam, Expert System Language Technology, HP, Morphologic, Ontotext, PEARSON – Longman, Polderland, Q Sphere, Quinary, SDL Multilingual services, Sinequa, Synthema, SYSTRAN, Telefonica Investigación y Desarrollo, TEMIS, THAMUS, XtraMind


ISO/TC 37 Terminology and other language

and content resources

Founded in 1936/re-established in 1951 Scope: Standardization of principles, methods and applications relating to

terminology and other language and content resources in the contexts of multilingual communication and cultural diversity – SC 1 Principles and Methods (chair: L.-J. Rousseau, Secr. Sweden)

– SC 2 Terminography and Lexicography (chair: G. Budin, Secr. Canada)

– SC 3 Computer applications (chair: B. Nistrup Madsen, Secr. Germany)

– SC 4 Language Resource Management (chair: L. Romary, Secr. Korea) Each SC has several working groups which run at least one project Based on practical needs horizontal cooperation and coordination is to be

guaranteed by SC chairs


Language Resource Management Standardization

Standardization is needed for language resources (mono- and multilingual), e.g. speech data, written (full) text corpora, lexical (general language) corpora and their processing methods

Relevant research areas are computational linguistics and computational lexicography, language engineering, etc., which have provided industrial best practices to be turned into official standards

This process will contribute to the further development of the language industries at large

As is the case with terminologies, language resources in general are often multilingual, multimedia and multimodal


ISO/TC 37/SC 1

The following standards are under the direct responsibility of ISO/TC 37/SC 1:

ISO 704:2000 Terminology work – Principles and methods ISO 860:1996 Terminology work – Harmonization of concepts and

terms ISO 1087-1:2000 Terminology work – Vocabulary – Part 1: Theory and

applicationThe following standards are under preparation: ISO/CD 704 Terminology work – Principles and methods ISO/CD 860 Terminology work – Harmonization of concepts and terms ISO/PWI 1087-1 Terminology work – Vocabulary – Part 1:

Theory and application ISO/WD 22134 Practical guide for socioterminology


ISO/TC 37/SC 2

Title: Terminography and lexicography Scope: Standardization of terminological and lexicographical working

methods, procedures, coding systems, workflows, and cultural diversity management, as well as related certification schemes


ISO/TC 37/SC 2 (2)The following standards are under the direct responsibility of ISO/TC 37/SC 2: ISO 639-1:2002 Codes for the representation of names of languages

– Part 1: Alpha-2 code ISO 639-2:1998 Codes for the representation of names of languages

– Part 2: Alpha-3 code ISO 1951:1997 Lexicographical symbols and typographical

conventions for use in terminography ISO 10241:1992 International terminology standards -- Preparation

and layout ISO 12199:2000 Alphabetical ordering of multilingual terminological

and lexicographical data represented in the Latin alphabet

ISO 12616:2002 Translation-oriented terminography ISO 15188:2001 Project management guidelines for terminology

standardization


ISO/TC 37/SC 2 (3)The following standards are under preparation: ISO/DIS 639-3 Codes for the representation of names of languages Part 3: Alpha-3

code for comprehensive coverage of languages ISO/CD 639-4 Codes for the representation of names of languages Part 4:

Implementation guidelines and general principles for language coding ISO/CD 639-5 Codes for the representation of names of languages Part 5: Alpha-3

code for language families and groups ISO/WD 639-6 Codes for the representation of names of languages Part 6:

Extension coding for language variation ISO/FDIS 1951 Presentation/representation of entries in dictionaries ISO/CD 10241-1 Terminological entries in standards – Part 1: General requirements ISO/CD 10241-2Terminological entries in standards ISO 12615 Bibliographic references and source identifiers for terminology ISO/NWI TR 22128 Quality assurance guidelines for terminology products ISO/NP 23185 Assessment and benchmarking of terminological holdings


ISO/TC 37/SC 3 (1)

title: Terminology management systems and content interoperability scope: Standardization of principles and requirements for semantic

interoperability, terminology and content management systems, and knowledge ordering tools


ISO/TC 37/SC 3 (2)

The following standards are under the direct responsibility of ISO/TC 37/SC 3:

ISO 1087-2:2000 Terminology work – Vocabulary – Part 2: Computer applications

ISO 6156:1987 Magnetic tape exchange format for (withdrawn)terminological/ lexicographical records

ISO 12200:1999 Computer applications in terminology – Machine-readable terminology interchange format (MARTIF) – Negotiated interchange

ISO 12620:1999 Computer applications in terminology – Data categories

ISO 16642:2003 Computer applications in terminology – Terminological markup framework


ISO/TC 37/SC 3 (3)

The following standards are under preparation: ISO/NWI TR 12618 Computational aids in terminology – Design,

implementation and use of terminology management systems

ISO/CD 12620 Computer applications in terminology – Data categories

NWI on conceptual modeling NWI on electronic catalogs


ISO/TC 37/SC 4 (1)

Title: Language resource management Scope: Standardization of specifications for computer-assisted language resource

management

linguistic infrastructures are being established or re-enforced as part of the rapidly evolving information and communication society;

professional activities involving language resource sharing and standardization are increasing in diverse areas: – governmental or non-governmental organizations, public or private institutions,

educational institutions, commercial enterprises, etc., – both, globalization and localization necessitate multilingual communication;

there is an increasing need for new standardization as well as urgent recognition of existing de facto standards and their transformation into International Standards


ISO/TC 37/SC 4 (2)The following standards are under preparation: ISO/NWI 21829 Terminology for language resources ISO/NP 23679-1 Word segmentation of written texts for mono-lingual

and multi-lingual information processing – Part 1: General principles and methods

ISO/NP 23679-2 Word segmentation of written texts for mono-lingual and multi-lingual information processing – Part 2: Word segmentation for Chinese, Japanese and Korean

ISO/CD 24610-3 Language resource management – Feature structures – Part 3: Word segmentation for other languages

ISO/WD 24611 Language resource management – Morpho-syntactic annotation framework

ISO/WD 24612 Language Resource Management – Linguistic Annotation Framework

ISO/WD 24613 Language resource management – Lexical markupframework


Approach

Industry requirements -> specifications for designing the standards Scoping: language technology standards, terminology and language

resource standards, etc. Feedback loops/co-operation schemes/liaison strategies/workshops Critically reviewing and revising existing standards to live up to changing

and new expectations and requirements


How to achieve terminological interoperability

Working principles– consensus building + international project management

– Conceptual precision by documenting and agreeing on DEFINITIONS (conceptual specifications)

– Linguistic precision

– Terminological assignments

– Systematic approach (concept systems, conceptual modeling)

– Multilingual, cross-cultural approach


Application example

CEN/ISSS WS ADNOM Administrative Nomenclatures


Goal: Implementing a framework for Semantic Interoperability across domains and languages in Europe

Without high quality and standards-based terminologies it is impossible to reach precision, efficiency, and transparency within and across eBusiness, eGovernment, eHealth, eLearning, eCulture, eScience, etc. processes and systems

Problem situation: lack of accessibility to high quality resources in many domains and languages, diversity of coding schemes and data organization -> lack of interoperability (syntactic, semantic, pragmatic) across existing data bases

Cultural differences across language communities and domain cultures are sometimes so strong that citizens as well as experts are lost in dynamic communication situations –> administrative and legal language is the best example


Goal: Implementing a framework for

Semantic Interoperability across domains and languages in Europe Such terminologies are organized and used for multiple

purposes in the form of dictionaries, data bases, thesauri, classification systems, nomenclatures, taxonomies, ontologies, indexes, term lists, etc.

ADNOM provides a coherent methodology for modeling, mapping, presenting, and accessing such resources in the context of the emerging European Interoperability Framework and in line with – and contributing to new – European and international standards (ISO/TC 37, JTC 1/SC 32, etc.).

Approach: federated registries based on ebXML and Topic Maps implementations, meta-data standards, terminology mark-up framework -> Semantic richness/complexity is managed and visualized and not eliminated.


Achievements – Results - Deliverables

Survey on existing administrative nomenclatures and similar terminological resources in Europe as well as on the organizations managing these resources

Procedural methodology described in the CWA (with principles and recommendations)

Demonstrator showing the implementation of the ADNOM approach (ebXML + Topic Maps (XTM) + ISO 16642 + ISO 11179 + other standards) (will be operational online in April), showing the “ADNOM Seamless Knowledge Core model” with federated registries, navigation services, etc.

Emerging organizational network of relevant institutions (stake holders), integrating existing networks; dissemination efforts (conference on 1st of December 2005 in Brussels on the Communicative Government organized by NL-Term together with CEN/ISSS WS-ADNOM)


A concrete example of mapping multilingual administrative

nomenclatures Simplified example on names of government ministries and

agencies in Austria, France, and Germany linked to the different scopes and responsibilities of these administrative units as far as pension schemes are concerned

The asymmetries are presented by a visualization of the conceptual map as implemented in a Topic Map linking the data described above

It includes the meta-data level using the COFOG classification (incl. the terms in 16 languages for “economic affairs”

In the framework of the “ADNOM Seamless Knowledge Core model” for organizing conceptual hierarchies


Terminological Semantics


Basic model: the Semiotic triangle

The semiotic triangle is a basic semantic model in philosophy of language and semiotics:

concept

object sign

Signs (linguistic signs, such as terms) refer to objects (e.g. objects of investigation) by conceptualizations

Note: the arrows point in two directions (onomasiological and semasiological reference, ontological reference)

cognition

communicationreality


Specific aspects of terminological semantics

Focus on– Conceptual content of domain texts and domain terms– Semiotic assignments between concepts and terms– Conceptual relations between terms– Disambiguation, conceptual change, synonymy, polysemy,

cultural diversity (equivalence), vagueness– Reference to objects of investigation

Methods– Conceptual logic, prototypes, various semantic theories– Definitions as explicit meaning statements– Semantic components (conceptual characteristics) are made

explicit, text semantics, semantic features– Referential statements on objects– Documentation (terminography), prescription (standards)


A functional-systemic model of terminological semantics

Semantic coherence in semantic systems, semantic reference Primary functions include cognition (concept formation) and

categorization (classifying objects into concept classes) as well as semiosis (assigning signs to concepts and concept classes) and representation

Secondary functions include systematic achievements in discourse such as precision and consistency in naming

Tertiary functions include translation, data exchange, technical communication, harmonization of terminologies, etc.

The arrows that go across all three functional levels signify the fact that semantic continuity and semantic interoperability depend on the fulfillment of multiple functions, in particular on the primary and secondary levels.


A poly-functional terminology model

Primäre Funktion

Terminology

Primary Functions

Secondary Functions

Tertiary Functions

Cognition, CategorizationSemiosis, Representation

Organization, Nomination

Consistency, Continuity, Precision, Understanding

Distinction, Recognition

Intercultural knowledge transfer, corporate Information Management, unambiguous communication, conceptual comparison of cultures

Learning and searching aids, Interoperability, Data exchange

Information, argumentation, knowledge sharing


Terminological reference

Due to the hypothetical, constructive nature of scientific knowledge, terminological reference is the process of constructing meaning hypotheses

This does not require the proof of existence of an object referred to

Terminological reference in social sciences and humanities typically includes both the postulation of existence of a certain object plus its conceptualization in a specific categorial schema, i.e. a particular theory

The theory is used as a categorial schema that allows the researcher to position objects or rather their linguistic representations in this schema


Formal models of terminological semantics and of semantic interoperability

The following models are examples of formal models that are created as instruments for managing semantic information and semantic structures and their representation forms

Semantic annotation has become a powerful tool for enriching the explanatory power of formal semantic models

Modeling tools in ISO TC 37:– ISO 11179 framework

– UML

– W3C standards/recommendations – XML, RDF(S), OWL, <<SKOS>>

– Conceptual modeling approaches, KR languages, etc.


Terminology Markup Framework (ISO 16642) – meta-model – structural skeleton


Lexical Markup Framework


LMF modeling approach

LMF models consist of UML classes, associations among the classes, and a set of ISO 12620 data categories that function as attribute-value pairs.

The data categories are used to adorn the UML diagrams that provide a high level view of the model.

LMF specifications, textual descriptions that describe the semantics of the modeling elements, provide more complete information about classes, relationships, and extensions than can be included in UML diagrams.


Semantic Interoperability Framework


Conclusions Terms are complex lexical units that are constantly semantically re-

determined by their communicative functions in professional discourse (incl. knowledge sharing) and by the semantic relationships to other terms in domain-specific knowledge structures, in text corpora, and in concrete action situations

A comprehensive, multi-dimensional modeling approach is suitable to describe and explain multiple functions of lex/term resources in diverse applications and situations

Modeling methods and tools adhere to a range of complementary standards that are constantly updated and improved (upward compatibility)

Modeling terminological semantics is a pre-requisite for reaching semantic interoperability in the Semantic Web and web-based information and communication systems

9 th open forum on metadata registries harmonization of terminology, ontology and metadata 20th –...

Documents

tools slide

xtramind slide

skos slide

open forum

metadata registry

iso standards

metalanguage iso

overview iso tc