МОКСЛО КАЛБОС КУЛЬТУРОС ИР ТЕРМИНОЛОГИЈОС ПРОБЛЕМОС

ЛИЕТУВИЮ КАЛБОТYРОС КЛАУСИМАІ, XXIV(1985)

IRINA NIKOLAEVNA VOLKOVA

TERM BANKS AND THEIR PRACTICAL USE

The scientific and technological revolution has increased the importance of scientific and technical terminology for the national economy. The creation and implementation of automated control and information systems, the transition to program-target planning methods, and the development of international cooperation and integration in addressing scientific, technical, and economic problems impose increasingly stringent requirements on the language of science and technology with regard to the unambiguity of terminology used in various types of documentation.

As a result, work on the harmonization and standardization of scientific and technical terminology at the national and international levels has intensified considerably, leading to the creation of a large number of terminology standards and recommendations (international, regional, national, corporate, etc.).

The Council for Mutual Economic Assistance (CMEA) has adopted 160 standards on terms and definitions; the International Organization for Standardization (ISO) has published 180 international terminology standards; and the International Electrotechnical Commission has issued about 60 chapters of the International Electrotechnical Vocabulary (IEV) and special terminology publications (which have the status of IEC standards).

National standardization organizations in 40 countries have adopted more than 15,000 standards on terms and definitions.

Considerable progress in the harmonization and standardization of terminology has been achieved in the Soviet Union, where more than 700 state and industry standards on terms and definitions are in force, and over 100 collections of recommended terms issued by the Committee on Scientific and Technical Terminology of the USSR Academy of Sciences have been published.

Standardized and recommended terminology collections number in the hundreds of thousands. A necessary condition for using these collections in practical and scientific work and for quickly finding the required terms is to present standardized and recommended terminology in the form of a particular organizational structure. One common way of structurally organizing large terminology collections is to create terminology banks

46

of data in which the storage, processing, and retrieval of terminological information are carried out using computers.

The term data banks usually refers to a collection of large, structurally complex datasets, together with the technical equipment and software needed to accumulate, process, and retrieve the required information.

A collection of datasets is commonly called a database. In terminology banks, the database consists of the terms they contain and the associated data elements. These collections are called terminological entries; the terms terminological record and dictionary entry are also sometimes used in this sense.

According to information published by the International Information Centre for Terminology (Infoterm), several dozen terminology banks are operating or under development around the world¹. Such banks exist in Great Britain, Venezuela, the GDR, Spain, Canada, Luxembourg, the USSR, the USA, France, the FRG, Sweden, and a number of other countries.

It should be noted that the organizations in which terminology banks are established, as well as the lexical collections that formed the basis for their creation, are quite diverse.

A number of terminology banks were created from terminological collections assembled during the period of active research into machine translation: the TEAM system of the linguistic service of Siemens (FRG), EURODICAUTOM (Luxembourg), the TERMIUM and TERMINOC systems (Canada), and others. TERMDOK, the system of the Swedish Centre for Technical Terminology, was created on the basis of the centre’s lexicographic service, which has been actively engaged in publishing terminological dictionaries for many years. The NORMATERM (France) and TERM (FRG) terminology banks were established within standardization services on the basis of collections of standardized terms.

The widespread creation and implementation of information-retrieval systems made it necessary to develop linguistic support for them; this was accompanied by the organization of collections of terms, which also played a role in the development of terminology banks.

Terminology banks are characterized by indicators such as the composition and size of the database (coverage of subject areas, the set of data elements in a terminological entry, and the number of languages in which terminological information is presented), the normative status of the terminological information, and user categories.

The composition of its database is of great importance to the functioning of a terminology bank. The greater the number of data elements …

1 World Guid to Terminological Activities (Organizations, Terminology Banks, Committees).— Infoterm Series A. München: Verlag Dokumentation, 1974.— 312 p.

47

the more terminology entries it contains, the greater the opportunities for users to use this bank.

For example, in the TEAM terminology bank, each terminology entry may contain up to 99 data elements (the developers of this bank call them categories). These include the main term, its synonyms, one or more definitions or a context, bibliographic information (source, author, publication date, etc.), grammatical parameters of the term, a description of its field of application, as well as equivalent terms in English, French, Spanish, Russian, Italian, Portuguese, and Dutch (equivalent terms may not be available in all languages).

Of the data elements listed above, about 25 are actually used for processing German terminology².

A terminology entry in the TERMDOK bank provides for a large number of both linguistic and extralinguistic data elements.

A terminology entry may contain up to 100 data elements relating to linguistic information for 26 different languages. At present, linguistic information covers 10 languages: Swedish, English, French, Spanish, German, Danish, Norwegian, Finnish, Russian, and Japanese. The 10 most frequently used linguistic data elements are: 1) main term; 2) synonyms and quasi-synonyms; 3) antonyms; 4) pronunciation, grammatical characteristics, abbreviations, definitions, usage labels, notes, cross-references ³.

Among the extralinguistic data elements, about 10 are most frequently used: document designation, term number, classification codes, date of adoption (publication), reference to an expert, and references to illustrations.

According to the degree of standardization of the terminology database, existing terminology banks can be divided into two groups: banks based on standardized terminology and functioning, as it were, as normative dictionaries, and banks compiled and supplemented not only from normative sources. For example, the terminology data bank of the French Association for Standardization (AFNOR), NORMATERM, includes terms established by French terminology standards; terms contained in the ter-

² Brinkmann K. H. Use of the TEAM Terminology Data Bank for the Terminology Work of the Deutsches Institut für Normung.— Infoterm Series 6. München—New York—London—Paris, 1981, p. 444.

³ The TERMDOK system.— TNC, 1978.— 45 p.

48

terminological appendices to industrial standards, as well as in international ISO standards on terms and definitions⁴

The TERM terminological data bank, which is being created within the framework of the interrelated data banks of the German Institute for Standardization, is also normative in nature, i.e., it includes only standardized terminology.

Such term banks as TEAM of Siemens, the Federal Office of Languages (FRG), TERMIUM and TERMINOC (Canada), TERMDOC (Sweden), and a number of others are not entirely normative and contain not only standardized terminology. Therefore, the information they provide is reliable (in terms of its documentary status), but does not always establish the official preference for using one term or another.

When compiling these banks’ databases, specialists “evaluate” each term and assign it a so-called “weight index.” For example, in the terminological data bank of the University of Montreal, a weight index of 0 means that the term is unacceptable for use; 1 means that its use is doubtful; 2 means that the term may be used; and 3 means that the term is standardized.

By language composition, the banks are classified as: monolingual, i.e., those in which all data elements of a terminological entry are presented in only one language (for example, the neologism bank in France contains terminological information only in French); partially bilingual or multilingual, i.e., those that provide equivalent terms (and sometimes definitions and a number of other terminological data elements) in several languages. TEAM is one such bank; in addition to the term and the data elements defining it in German, its database also includes terms (and sometimes definitions) in English, French, Spanish, Italian, Russian, and Dutch. The EURODICAUTOM bank includes terminology in English, German, Danish, French, Italian, and Dutch; term banks are fully bilingual. Such banks include those in Canada, which, owing to demographic characteristics, are entirely English-French, i.e., each terminological entry in their database consists of two identical blocks: one in English and one in French.

The aim of many term banks is to assist, first and foremost, translators. These include TEAM of the Linguistic Service of Siemens (FRG), EURODICAUTOM of the European Economic Community (Luxembourg), and TERMIUM of the University of Montreal (currently the official term bank

⁴ Laurent J. La banque automatisée de données terminologiques de l’AFNOR NORMATERM.— Courier de la Normalisation, 1979, N 245, p. 46—51.

4 Užsak. Nr. 1007

49

of the Government of Canada), TERMINOC—the term bank of the province of Quebec (Canada), and TARGET—the term bank of Carnegie Mellon University (USA).

One of the tasks of the term bank for technical terminology being established in Great Britain is to monitor the correct use of English terms included in existing term banks (for example, those of Sweden and the FRG).

The tasks addressed with the help of Siemens’ TEAM bank also include lexicographic work—the preparation and publication of various dictionaries. A DIGISET phototypesetting machine is used for this purpose.

The technological process for the automated preparation of dictionaries can be represented as follows: selecting a specific set of terms from an automated terminological data base; sorting and presenting the data in the format customary for dictionaries; preparing a typesetting page (i.e., a frame of photographic film containing text laid out in formats customary for dictionaries); and printing.

The advantages of machine-produced dictionaries include the space-efficient arrangement of accumulated information; the ability to update the data collection quickly and, consequently, to prepare dictionaries that reflect the current state of the art with maximum timeliness; and the prompt preparation of dictionaries with different sets of data elements.

There are several term banks in the Soviet Union. One of them operates at the All-Union Scientific Research Institute for Technical Information, Classification and Coding (VNIiki) of the State Standards Committee: the Automated System for Information and Terminological Services (ASITO), which accumulates data on standardized terminology. The system’s main tasks are to increase the effectiveness of implementing standardized terms, improve the timeliness and quality of information and terminological support for the national economy, promote coordination among terminologies under development and standardization, and fulfill the USSR’s obligations within ISO, IEC, OIML, CMEA, and other international organizations in preparing Russian versions of international terminological standards and dictionaries.

ASITO was developed on the basis of a collection of terminological documents and the system of manual bibliographic and terminological cards of the VNIiki Reference Term Bank (RTB). ASITO is an evolving system. The first stage, ASITO-1, has been in industrial operation since 1981; at the same time, work is under way to improve and modernize the system by creating ASITO-2. ASITO-2 is planned to expand the range of documents and data elements processed, implement new types of searches, for example by definitions, and expand the set of output formats, etc.

50

ASITO-1 processes data from state standards, CMEA and ISO standards on terms and definitions. As in other automated terminology systems, the basic storage unit is a terminology record containing a set of data elements with terminological and bibliographic information about a specific standardized term.

In ASITO-1, each terminology record includes a heading—the standardized term; acceptable and unacceptable synonyms of the standardized term; a short form (for multiword terms); equivalents of the standardized term in English, German, and French; a definition of the concept; letter symbols for quantities; the designation and title of the standard that established the term; the term number in the standard; the date on which the standard was approved and entered into force; and the standard’s classification code. In all, the ASITO-1 automated database contains about 70,000 terminology records.

The software supports database maintenance and various types of search. The primary type is interactive search using a display. Searches can be conducted by any data element in a terminology record, or by combinations of elements. The system supports searches that take into account inflectional variants of words and syntactic variants of multiword expressions. It also supports searches and retrieval of information when the query does not lexically match the term entered in the system exactly, as well as truncation of the ends of query word forms. Thus, if the bank contains the term discrete random sequence, information about it should also be retrieved if the query uses a different order of term elements: random discrete sequence. Likewise, a different ending in the query term from that of the term in ASITO should not affect the completeness of the response. Thus, the system allows queries to be entered in natural language without prior indexing, makes it possible to restrict searches to a specific subject area, and allows users to specify the category of normative and technical documents and other parameters.

The system can compile various indexes: consolidated and systematic, simple and permuted (KWOC-type), by category of normative and technical documents, by Russian-language terms and foreign-language equivalents, and so on.

Information can be exchanged on magnetic tapes using a unified communications format for terminology data. Its structure is harmonized with the communications format for bibliographic data in accordance with the international standard ISO 2709, while its content conforms to the draft format

51

MATER (ISO 6156). Users of the system can obtain on magnetic tape either the entire body of standardized terminology or a portion of it in accordance with their query.

The ASITO terminology bank is used to provide terminology and information services to research and design organizations, publishing houses, and industrial enterprises (ASITO currently serves 400 collective subscribers); to exchange information with terminology banks both within the country and abroad; and to prepare statistical and analytical data on standardized terminology.

Given that the number of terminology banks will undoubtedly grow, a targeted, comprehensive program should be developed to create a nationwide system of terminology banks. Alongside banks of standardized terms, this system should also include banks of nonstandardized lexical units.

The system of terminology banks should be capable of functioning as an international terminology service that guarantees the accuracy and rigor of term usage in all documents across research, administrative and managerial, social and historical, and cultural and historical contexts; covers all branches of science and technology; and includes Russian terms and foreign-language (and national-language) equivalents used in the terminology systems of the national languages of the Union republics and in international professional communication.

A nationwide system comprising all terminology banks should apparently include a bank (or banks) of terms in the fundamental sciences, banks of general technical and specialized terms, and banks of terms in the national languages of the Union republics.

The existence, even today, of a large number of terminology banks in which terminology information is stored, processed, and retrieved using computers has raised the problem of their interoperability and of organizing the exchange of terminology data among them. The creation of an international network of interconnected terminology banks—the TERMNET system—is one of Infoterm’s tasks. To achieve this, an international communications format for exchanging terminology information on magnetic media is being developed. This work has been carried out for a number of years by Working Group 4 (WG 4), “Use of Computer Technology in Terminology and Lexicography,” established specifically within ISO/TC 37, “Terminology. Principles and Coordination.” This working group has prepared a draft international standard 6156, “Format for the Exchange of Terminological Data on Magnetic Tape” (abbreviated as the MATER format). The purpose of developing this standard is to establish requirements for the format and for

52

alongside the storage of data on magnetic media. It is intended for the exchange of monolingual and multilingual terminological information⁵.

The widespread prevalence of terminology banks is significant. This article offers only a very general overview of terminology banks and of the World Terminology Network, which is currently being created.

As already noted, interest in terminology banks is shown abroad not only by standardization services and information centers, but also by government agencies and large companies. Substantial resources are being invested in the creation of terminology banks.

This is because terminologies record the current level of scientific and technical knowledge. The further development of the sciences, scientific communication, matters of government planning and statistical accounting, teaching, and information activities are impossible without the adoption of terminologies, nomenclatures, product marks, letter symbols for quantities, and various abbreviations that are uniform and understood in the same way by everyone.

⁵ I. P. Perstnev, I. N. Volkova. A communicative format for exchanging terminological data on magnetic tape.—Scientific and Technical Terminology, 1980, no. 9, pp. 22–24.