Section on Classification and Indexing (original) (raw)

IFLANET home - International Federation of Library Associations and InstitutionsActivities and ServicesSearchContacts

Newsletter of the Section on Classification and Indexing

No. 18

December 1998

To all members of the Section

It was a great pleasure to meet so many colleagues and friends at the Amsterdam meeting. Our programmes were well attended and the speakers were stimulating and informative. As you may recall, we decided that in view of the large number of very useful reports that we received from committee members, it would be a good idea to issue a second number of our Newsletter this year. I have therefore prepared one and have pleasure in circulating it now, with my best greetings for the festive season.

Ia McIlwaine
Chairperson
IFLA Section on Classification and Indexing

Standing Committee of the Section on Classification and Indexing

The Standing Committee now has its full complement of 20 members. These are: Marje Aasmets (Estonian Academic Library, Tallinn, Estonia), Jon Anjer (Faculty of Journalism, Library and Information Science, Oslo College, Oslo, Norway), Pilar Benedito Castellote, (Biblioteca Naçional, Madrid, Spain), Lois Mai Chan (School of Library and Information Science, Lexington, Kentucky), Michel Fournier (Université Laval, Québec, Canada), Friedrich Geisselmann (Universitätsbibliothek Regensburg, Germany), Elisa Grignani (Università degli Studi di Parma, Italy), Magda Heiner-Freiling (Die Deutsche Bibliothek, Frankfurt am Main, Germany), Adriana Király (Biblioteca Judeteana "Octavian Goga" Cluj, Romania), Pia Leth (Kungl. Biblioteket, Stockholm, Sweden), Elisabet Lindkvist Michailaki (The Swedish Parliament, Stockholm, Sweden), Ia McIlwaine (University College London, UK), Max Naudi (Bibliothèque nationale de France), Gerhard Riesthuis (University of Amsterdam, Netherlands), Poori Sultani (National Library of the Islamic Republic of Iran, Tehran, Iran), Edward Swanson (Minnesota Historical Society, Saint Paul, USA), Marie-Martine Tomich (Bibliothèque de l'Université René Descartes, Paris, France), Irina Tsvetkova (National Library of Russia, St Petersburg, Russia), Júlio Vaz dos Santos Rodrigues, Fundação Calouste Gulbenkian, Lisbon, Portugal), Marcia Zeng (Kent State University, USA).
The Standing Committee has one corresponding member: Mandana Sadigh-Behzadi (National Library of Iran, Tehran, Iran) and one observer: Marie-France Plassard (IFLA UBCIM Programme, Die Deutsche Bibliothek, Frankfurt am Main, Germany)

You will recall that 1999 is an election year, and all present standing committee members have been circulated with the appropriate form. Other persons seeking election should ensure that they are proposed by their institution and another one, unless they are nominated by the National Member, and if anyone cannot find a suitable proposer, please inform the Chair or the Secretary, who will endeavour to assist. One of the conditions of joining the Committee is that candidates should be in a position to attend at least 3 out of the 4 IFLA meetings covered by their period of membership, with the proviso that IFLA is unable to make any contribution towards expenses.

News from OCLC

In September it was announced that Joan Mitchell, chief editor of the Dewey Decimal Classification since 1993, with succeed Peter Paulson as executive director and editor-in-chief of OCLC Forest Press. Mr Paulson will retire on December 31st 1998 and in the meantime Joan has been acting as deputy director. In conjunction with the management change, by June 1999 OCLC Forest Press offices will be relocated from Albany, New York, to OCLC's headquarters in Dublin, Ohio. I am sure that everyone will wish to congratulate Joan on her new position, and to wish Peter a long and happy retirement.

Developments in classification and indexing in Estonia, 1997/98

The main problems in the field of classification have been connected with the UDC. The translation into Estonian and the editing of the translation/selection were finished by the summer of 1997. Since then the editors have spent a lot of time in reading proofs of the UDC. Unfortunately the printing of the tables has been delayed, mainly due to financial problems. That is why the Estonian UDC has not been published yet.

The essential problems of classification and indexing have been discussed at regular meetings of the Working Group (WG) of Classification and Subject Indexing of the Estonian Librarians' Association once a month. The members of WG are mainly representatives of academic/research libraries, but WG has standing contacts with larger public libraries as well.

The application of the two principal systems - the UDC and the list of subject headings, is swiftly spreading in Estonian libraries. While public libraries as "newcomers" have less experience in working with those systems, academic/research libraries are extremely busy with unifying the use of the UDC in Estonia and keeping up with its revisions.

At the meetings of the WG the main discussion points in connection with the UDC have been the application of common auxiliaries and the complicated systematization of materials/ themes in Class 8 Language and Literature. The practical aspects of using all common auxiliaries have been discussed. As a matter of fact, Tables 1k-03 and 1k-05 are not used as often as they should be. Various dissimilarities have been revealed in the practical use of common auxiliaries of form by different Estonian libraries. The WG has worked out (in connection with that) several recommendations and distributed them among the libraries. The practical application of the common auxiliaries of point of view has been discussed thoroughly, especially in view of the article of G. Robinson.

The new structure of class 8 is not used in all Estonian libraries, so many questions about systematizing different themes have been under discussion, namely various types of dictionaries, history and theory of literature and literary criticism, old linguistic texts, fiction written in a dialect, etc. Several recommendations of the WG have been subsequently disseminated among the libraries.

The main features of the UDC, especially the changes in the classification, have been repeatedly treated in lectures, seminars and courses. The Estonian Academic Library organized a two-day course of lectures. Lectures and seminars have been held mainly for public and school libraries.

The work on a universal standardized list of subject headings (thesaurus) is going on. (By the end of 1997 the thesaurus consisted of 17,557 terms on 65 themes).

The text of the "Principles Underlying Subject Heading Languages (SHLs)" has been distributed among libraries and later discussed by the WG.

The history and present day practical situation of subject indexing in the Estonian Medical Library (founded 1944) was discussed at a meeting of the WG in the autumn. Up to 1991 the alphabetical subject catalogue of the library was held in Russian. Since 1992 subject indexing has been done in Estonian (with parallel descriptors from MeSH in English).

Time and again the WG has discussed how to use different fields of INNOPAC, e.g. the field for topical subject headings of the Estonian Universal SHL would be 650, fields for special (narrower) subject headings of libraries would be 690-698, etc.

Our problems continue to be manifold. We are still in need of professionals and manuals (especially in Estonian). Great efforts have been made to translate/create the UDC tables and list of subject headings in Estonian. This is a long-lasting process that requires the co-operation of all Estonian libraries.

Marje Aasmets
Estonian Academic Library

Subject indexing and classification developments in France, 1997-1998

With regard to classification, the major event this year is, of course, the translation of the 21st edition of the Dewey Decimal Classification; firstly with the publication of an abridged edition (Éditions Cercle de la librairie), and especially with the launching of the complete edition in French, the first for 25 years, which will take place during this IFLA congress.

As the result of a partnership between Canada (ASTED, BNC, BNQ) and France (BnF), this joint edition is a significant experiment, which will be reported in our second workshop (Thursday, August 20th, in the afternoon). It gives us a global view of DDC, thus enabling us better to appreciate what seem the principal problems and to formulate some suggestions.

From the national point of view, DDC is currently used principally to establish call-numbers and for locating documents, often with many adaptations which the majority of users find essential, and not in any standardized form. On the other hand, as is the tradition in France for any classification, it plays very little part in subject access.

At present in the BnF only the collections on open-access are indexed by Dewey and the numbers provided are translated into words to allow searching and form the basis of the development work of the Dewey BnF authority file. Our experiences during the work of revision have refined our thoughts on its future use which will probably be directed towards less specific Dewey numbers and will certainly have greater conformity with the national subject headings system RAMEAU. Finally, within the Permanent UNIMARC Committee (PUC) the institution also participates in the development of the UNIMARC Format for Classification Data.

Still on the subject of formats, the BnF in 1998 developed, in collaboration with other libraries, a French interchange format for authority records in UNIMARC.

The year 1997-98 has also been particularly significant with regard to formats and systems. In university libraries the project "Système universitaire de documentation (SUD)" [University system of documentation] was established at the beginning of 1997 with ABES (Bibliographic Agency of Higher Education). It should become operational in 2000, and it will integrate in UNIMARC all the catalogues of existing networks and will be the sole system of all French university libraries.

At the BnF the data of all existing catalogues and databases have been converted into the new INTERMARC Integrated Format, which gives a new structure to the BnF records, in particular, and has migrated them to the new Système d'information (SI) [Information system]. With more than ten million online records, the SI will be operational and able to be consulted when the rez-de-jardin reading rooms open next October. It should be fully operational and in production by the end of 1999.

With the SI, as with SUD, the standard for subject access is RAMEAU, the national subject heading language, initiated in 1980 and developed for networking since 1987. From 1999, in conjunction with the new format of the SI, its management in the BnF will be more focused on proper "subject headings" (common nouns) since proper names (authors, corporate bodies, titles) controlled by the appropriate authority files, will have a single form of authority and will be usable as a subject as well as an author point of access.

RAMEAU is used by all types of libraries for all types of document, and is presented and described in the IFLA document on "Principles underlying SHLs" where the section relating to it has been thoroughly reviewed and updated to the end of 1997/beginning of 1998, within the Working Group of our Section.

New initiatives for the development and management of our SH Language are taking place:

At the beginning of 1998 a new RAMEAU tool was already developed on the BnF's authorities CD-ROM, with an English-French index allowing direct search by LC subject headings equivalent to the RAMEAU ones, and a word index allowing search by the words of these English subject headings.

These concordances and the experience gained since the beginning of the nineties were of great assistance for the "multilingual European thesaurus" project which will be presented during the Open Forum of our Division (Monday, August 17th, Magda Heiner-Freiling).
This co-operation between European national libraries, which has been very active since the end of 1997 within a COBRA+ Working Group (BL, BnF, BNS, DB) is seeking to exploit the richness and the flexibility of national subject headings languages, to allow multilingual subject access to bibliographical databases through the establishment of links of equivalence between the headings of the subject systems used.

Lastly, in conclusion on the note of co-operation, it should be noted that RAMEAU also continues to co-operate actively with other countries in Europe (Polish libraries and the thesaurus KABA, public libraries of the French community of Belgium) as in other French-speaking countries (the envisaged renewal at the end of 1998 of the agreement with the Bibliothèque de l'Université Laval, Répertoire de vedettes-matière).

Max Naudi
Bibliothèque nationale de France

Report from Germany

The Project MUSE - a multilingual thesaurus project of European National Libraries

In spring 1997 a cooperation project between four European national libraries started with the aim to establish links between the different national authority files for subject headings. The initiative resulted from the efforts of the Schweizerische Landesbibliothek Bern with its collections in French as well as in German - for French publications the French subject authority file RAMEAU is used for indexing purposes in the Bibliographie de la France, whereas German publications are indexed with subject headings taken from the German Schlagwortnormdatei (SWD) in the Deutsche Nationalbibliographie. Besides, there is a strong interest in both German and French speaking countries in the Library of Congress Subject Headings (LCSH) which are not only used for the indexing of American literature, but also lately by the British Library for the British National Bibliography and by many other national bibliographies around the world. After first contacts with some other European countries (among them Italy and Belgium) a working group was established with two members each from the British Library, the Schweizerische Landesbibliothek, the Bibliothèque nationale de France and the Deutsche Bibliothek.

Two working group meetings in December 1997 in Paris and in April 1998 in London and a number of informal contacts and constant e-mail connections have led to first positive results and a working plan for the next months. Two different and rather specific fields of knowledge were selected to extract all relevant subject headings from the three authority files (LCSH, RAMEAU and SWD) and to establish links between the descriptors: sports and theatre. Though the systematic approaches especially between LCSH and SWD are rather far from each other (broad precoordinated subject headings in LCSH vs. specific subject headings with postcoordinated strings in SWD) it was possible in about 70 % of circa 300 descriptors in each subject and authority file to establish links between the systems. At present we try to solve the difficult cases as far as specific questions in the subjects are concerned and discuss the methodological questions of the approach in general. An interim report of the project will be discussed during an informal meeting in Amsterdam and at our next regular session in Bern in the middle of September. Among the next steps in the project the establishment of a powerful electronic connection between the different national databases for subject headings certainly will be the most important, but also most difficult to realize.

The project was named MUSE (MUltilingual Subject Entry) to make clear that it does not want to translate terms from one language to another or establish a homogenous thesaurus with terms in different languages. The aim is the linkage between already existing national thesauri which remain in their own linguistic surroundings, but offer the opportunity of switching to the user of an OPAC or a national bibliographic database. Without even noticing the changing language access the user should be able to extend the search for literature to publications in other databases. By the end of 1998 the pilot project should be finished with a trilingual list of descriptors in the fields of sports and theatre and the interim report. Additionally the working group selected a number of publications by internationally well-known publishers which were indexed in all three systems and will compare the quality and congruence of the subject heading strings used for the same title in LCSH, RAMEAU and SWD.

The partners will try to seek funding by national sponsors (as the DFG in Germany) during the first stage of the project.

As soon as the methodological and data processing problems are solved (during 1999) the project will be extended to other subject fields as well. The integration of additional languages (with priority to Italian and Dutch) is possible, but of course it depends on the cooperation of the national libraries or bibliographic agencies. The integration of multilingual thesaurus activities by special libraries/documentation centres and their national and international networks is very desirable, too, and first contacts have been established for instance in the multilingual ETHERELI project by theological libraries in different European countries which is concerned about the development of a thesaurus in the field of theological ethics. There is also a strong interest in cooperation with the project in a number of national networks of special libraries as, for instance, in Germany among the art, museum and theatre libraries.

For the growing international relations a multilingual thesaurus would also be very desirable, but we also need their professional advice when establishing the links between very specific and not easily to translate descriptors in the different subject heading systems. And sometimes even terms which seem to be identical turn out to be completely different - "Schwarzes Theater" in the German authority file SWD as something absolutely different from "Black theater" in LCSH and "Théâtre noir" in RAMEAU. We are aware of the difficulties in detail as well as in the project in general (especially as far as data processing is concerned), and it is certainly too early to say that it will be a success after all. In fields of knowledge which are closely related to the cultural, historical, linguistic and administrative specifics of a country, e.g. in the humanities and partly in the social sciences (in the first step represented by theatre) we will have less equivalent relations between the three systems than in science and technology or a field like sports. Yet the first results are encouraging enough to proceed with the work.

New initiatives in classification

During recent years classification has not been one of the main issues in the discussions of German academic librarians. The progress in verbal indexing, marked by the development of a widely accepted manual (RSWK) and a German subject authority file (SWD), was impressive. But the online environment and the new possibilities for subject queries, offered by the Internet, showed also the necessity of a widely accepted classification as an additional systematic approach, broader than the extremely specific subject heading strings of RSWK. An expert's report on different national and international classifications will be published this Autumn by the Deutsche Bibliotheksinstitut, accompanied by a pamphlet with the programmatic title "Back to classification!". Contacts within the IFLA Section on Classification and Indexing convinced the German members of the Standing Committee of the benefits of an internationally accepted and translated classification as DDC, 21st ed. and the comfortable handling of DEWEY for WINDOWS. They decided to present it to those German librarians who are responsible for decisions in the field of subject indexing and classification. Joan Mitchell, the editor of DDC was invited for this workshop which will be held in Die Deutsche Bibliothek Frankfurt in October and may, as we hope, lead to an initiative for a German translation of DDC.

Magda Heiner-Freiling
Die Deutsche Bibliothek

Report from Iran

1. Classification

1.1 Iran has developed and expanded all sections relating to Iran in both DDC and LC classifications; that is to say, history, literature, language, geography and religion. For all of these the Library of Congress has been consulted. All of them have reached a second, and some a third or fourth edition. The newest is the long-awaited 2nd edition of Islam in the LC classification, which has been due for some years. Ms. Zorah Alavi edited and it came out early in 1998 (although the title page states 1997). Since all of them are used in the National Library of Iran and both DDC and LC appear on our catalogue cards, additions and changes are included each time a new impression is needed.

What we are still working on is the section of Iranian/Islamic philosophy in the Library of Congress Classification. We hope that it will be finished by the end of this year and hopefully published next year. Much research has been done for each minute detail.

2. Translation

2.1 The Abridged edition of DDC 12th edition was translated into Persian and is now widely used in most public and school libraries. It includes the abridgement of our expansions mentioned above, so that it can meet the needs of our libraries. The methodology we used was translating and expanding where necessary and omitting superfluous detail from both the English and Persian introductions.

2.2 In using the Library of Congress Classification for Persian matters, we realized that in many cases we have to translate the whole section. For example, in literature the LC schedule has listed individual authors by century and in each century in alphabetical order. The order of the Persian alphabet is different from that of the Roman. So, if we want to be loyal to the classification system, we have to rewrite the names in Persian and arrange them according to our alphabetical order. Of course, things are not so simple. This was the subject I proposed to talk about at an IFLA meeting.
We have already done some more urgent sections, such as PQ: French Literature (18th, 19th, 20th century); PR: English Literature, and so on. In many other sections, wherever LC says "Arrange alphabetically by subject, etc." we have to arrange them according to our alphabet, and hence translation is needed.

3. Subject Indexing

3.1 The National Library of Iran publishes Lists of Persian Subject Headings which is now used in almost all libraries in Iran, with the exception of some medical libraries which use MeSH.
The Persian Subject Headings has some supplements. The latest, full, 2nd edition came out in 1994. It is kept on computer as well, and a CD-ROM edition is published too. The third edition is in press, we are in the process of proof-reading and inserting the latest additions into it.

3.2 Persian Medical Subject Headings based on MeSH came out in 1993, by Ms. F. Rahadust, at that time head of the Library Group at the Iran University of the Medical Sciences. This has been completely revised and is being published under the title Persian Medical Thesaurus.

3.3 Another interesting thesaurus is the Persian Cultural Thesaurus which came out in 1996. A group of specialists worked on it under the supervision of Dr. F. Khosravi. It is also computerized and gives related, broader and narrower terms. It is based on the ISO Guidelines for the establishment and development of monolingual thesauri. 2nd revised edition.

4 Indexing

A project was started last year in the National Library of Iran under the title "Indexing and Abstracting Services". In order to reach this goal the following steps are foreseen:

  1. To develop "Standards for indexing". For this purpose a committee has been assigned.
  2. To gain some practical experience, all the MLS and PhD theses in library and information science of Iranian students are being indexed and abstracted. This work is finished and is kept on a computer as a database. A book format is planned to be published. This was done under the supervision of Dr. M. Sadiq-Behzadi. We are now beginning to work on the Standards of Indexing.

Poori Soltani
National Library of the Islamic Republic of Iran

Report from Norway about the field of classification and indexing

Since summer 1997 the main activities have been planning, in this order of priority: a. new Norwegian abridged Dewey Decimal Classification, based on DDC 21
b. Norwegian updated guidelines for indexing

  1. Several of the National Library's functions have up to now been taken care of by the University Library in Oslo (UBO), and this part of UBO will become a National Library department in Oslo at the beginning of 1999. It is planned that this Department will have the national responsibility for Norwegian standardization in the fields of Cataloguing, Classification and Indexing (most important issues Cataloguing Rules, MARC, Authority files, Norwegian DDC editions, Guidelines for indexing). The Norwegian Committee on Classification and Indexing will be appointed by the National Library from next year (now there are two appointing agencies, one from the field of public libraries, the other from the field of university and special libraries).
    This implies two important decisions:
    1. The work with a Norwegian abridged translation of DDC 21 will be organized by the National Library (during the IFLA '97 Meeting the Norwegian Library Association was the planning agency for a new edition, meeting Forest Press in Copenhagen). We hope that the work will start late in 1998 or early in 1999. The National Library and OCLC/Forest Press will conduct a meeting in Amsterdam in connection with the IFLA conference.
    2. We are discussing making updated Norwegian guidelines for indexing. The Norwegian Committee on Classification and Indexing has proposed this to the National Library as a limited project, and we hope that this can be done in 1999. The work "Emneordskatalogisering: forlag til norsk standard" (Norwegian proposal for guidelines for the constructing and maintenance of an alphabetic subject catalogue) by Ellen Hjortsaeter will be a good starting point, but as this work was written in 1990, updating will be necessary, especially to show relationships between terms.
  2. Oslo College, in co-operation with the Norwegian Committee on Classification and Indexing, held a seminar in May, (6th-8th) on "Verbal subject entries to OPACs". One of our main speakers was Dr A. Steven Pollitt, CeDAR, Huddersfield University, with "Faceted classification as Pre-coordinated subject indexing: multi-dimensional searching for OPAC users". His lectures included a very interesting part "Propects for using Dewey Classification in a View-based Searching OPAC Dewey Decimal Classification". More than a hundred delegates met at the seminar, among them both the Standing Committee members from Sweden. The programme (English translation) is available at http://info.rbt.no/nkki/korg98/prog-eng.htm. Reports and/or html coded slides can be found on http://info.rtb.no/nkki/korg98/referat.htm, all of them apart from Pollitt's slides, in Norwegian.

Jon Anjer
Oslo College

Developments in Portugal regarding Classification and Indexing

The indexing system in Portuguese - SIPORBASE- was first published by the National Library in 1980; a second edition, revised and enlarged, was published in 1992. Since then the manual has been occasionally updated and a third edition is presently in the press, hopefully to be available by the end of this year (1998). The main changes consist of the inclusion of a new chapter of instructions intended for the analysis and indexing of special documents - concerned with their physical and/or bibliographical features. In the previous edition of SIPORBASE these instructions were considered together with those established for special subjects. A noteworthy addition is an alphabetical index which should certainly make the manual easier to use.

With regard to indexing languages, the CLIP project - harmonization of indexing languages in Portuguese - began in 1989 and is a project between the librarians of special libraries and the National Library to improve and harmonize the languages in use by the libraries for each subject area.

Several controlled vocabularies have been published, namely for Photography (1993), Design (1996), Military Architecture (1996) and Religious Architecture (1997). Presently, an edition for Illuminated Manuscripts is being prepared.

Classification

The UDC manual used by the National Library and by the libraries integrating their records into the National Bibliographic Database - Porbase, is a Portuguese edition of 1990, whose statement of responsibility belongs to the National Library. This manual has been regularly updated and a new revision is now being published and comprises changes to class 9 (History) as well as the inclusion of specific notations in class O for classifying Computer Science.

Júlio Vaz
Biblioteca de Arte, Fundação Calouste Gulbenkian, Lisboa, Portugal

Developments in classification and indexing: activities in Romania

The principal event in Romania in the past year was the publication of the first UDC International Medium Edition in Romanian. This was published in two parts, in December 1997 and August 1998 (Pubn. no. UDC - PO24, authorized by the UDC Consortium with licence no. UDC 9704). The selection of terms, translation and publication were carried out by the National Library of Romania and the Editor in chief was Georgeta Clinca. The translation is based on the Master Reference File and includes amendments up to Extensions and corrections to the UDC 18, 1996.

The importance of this event is increased by the fact that the UDC is the classification system adopted at present in all Romanian libraries. It will also promote the use of the same edition of the classification (from 1999) in national, public and university libraries, especially the new revision for class 8.

Prior to 1990, with the exception of some specialized scientific libraries which used co-ordinate indexing, Romanian libraries did not use alphabetical subject catalogues. Subject access was simply systematic, through the use of UDC class numbers and, as an additional aid, the alphabetical subject index of the UDC.

With the introduction and development of automation in the field of cataloguing, classification and information retrieval in more and more libraries, new trends in indexing have begun since 1994. The National Library of Romania and the public libraries provided, and continue to provide, access primarily by keywords from titles. The university libraries created co-operatively a minicontrolled vocabulary of Romanian keywords (applying uniform heading, synonymy and homonymy principles).

In 1996, on the initiative and under the co-ordination of the Bucharest Central University Library, the university libraries initiated a national programme for the construction of a Romanian subject heading system, based on the UDC structure, the alphabetical subject index of the UDC and especially using Library of Congress Subject Headings as a model.

In 1997 A. Andrian from the National Institute of Information and Documentation created a Thesaurus for science and technology (according to STAS 10711, ISO 8777 and ISO 5964) based on UDC and useful particularly for specialized libraries with 60,000 or more items. He is also preparing a Universal Thesaurus, again based on UDC, and intended for libraries with an encyclopaedic profile, and a stock of up to 150,000 items.

In order to communicate better and to share experience and work in the field of classification and indexing activities at national level, at the meeting organized in December 1997 by the "Octavian Goga" County Library of Cluj and IME Romania (TINLIB being used in most Romanian libraries) the National Commission for UDC was created. The elected members represent all types of Romanian libraries. The main goals of the Commission are:

  1. Evaluation and analysis of traditional and automated instruments used in classification and indexing of documents in Romanian libraries
  2. Use of the same UDC version (especially classes 004 Computer science, 8 and 9 and particularly, joining efforts with professionals from all types of libraries (including the National Library and some major public libraries) for creating a Romanian Subject Heading List

Other events in connection with the classification and indexing activities in Romania which have occurred since my last written report (1995) include:

Adriana Kiraly
"Octavian Goga" County Library
Cluj, Romania

Report on classification activities in Sweden

A. Royal Library

This report gives a brief outline of the issue of classification, indexing and subject headings at the Royal Library. But first a few words about the Royal Library and the Swedish national union catalogue - LIBRIS.

The Royal Library

The Royal Library is the National Library of Sweden. Its mission is to preserve Sweden�s cultural heritage and to support and promote Swedish science and research. Another important task is to promote library co-operation both nationally and internationally.
See also: www.kb.se/eng/kbstart.htm

LIBRIS database

KB through its LIBRIS Department manages the LIBRIS system, which is the Union Catalogue of Swedish libraries. The database of LIBRIS includes holdings of about 160 libraries, mainly research libraries, but also some public libraries. The database contains approximately 6 million titles. One million titles are Swedish and include the Swedish National Bibliography.

The LIBRIS system dates back to the 1970s. It is presently undergoing a major transformation from a mainframe environment to a client-server system. The NEW LIBRIS Project is expected to be realized during 1999.
LIBRIS WebSearch is available on the Internet,
see: www.libris.kb.se/english/home.html

The Division of Bibliographic Development and Co-ordination (BUS)

The Division of Bibliographic Development and Co-ordination (BUS) was established in connection with the reorganization of The Royal Library two years ago. The activity of BUS is very much focused on the bibliographic quality of the LIBRIS database. Some important tasks and projects of BUS are education in cataloguing in LIBRIS, guidelines for authority control and subject headings, retroconversion and digitizing unique material from the collections of the Royal Library and making them available on the Internet.

Classification

The classification system used at the Royal Library as well as at most university libraries and all public libraries is called Klassifikationssystem för svenska bibliotek. It is entirely a domestic system used nowhere but in Sweden.
The responsibility for the updating of this Swedish system is handled by a committee within Sveriges allmänna biblioteksförening, SAB, (= General Swedish Library Association). The latest edition was published in 1997.
The SAB classification system works quite well with more general collections whereas special libraries in technology and medicine prefer using UDC, Dewey and MeSH Terms.

Subject headings

As being part of the National Library BUS has the responsibility for the quality of cataloguing in our own library as well as in the LIBRIS database as a whole.
At the Royal Library there has been a long tradition of creating subject headings, starting from the 1920s, but with a down period from the 1950s to the 1980s.
One of our tasks at BUS has been to train cataloguers at the Royal Library to make subject headings of high quality.
We are very pragmatic but we try to follow general principles stated by IFLA. We use precoordinated strings but feel that we have no strict rules to follow.
Last autumn The Royal Library published guidelines called Ämnesordsindexering - en handledning (=Subject headings indexing: a guideline) by Unn Hellsten and Margareta Rosfelt. This publication is of some help but is at the same time too general for being of practical use.
Therefore we are now working on new guidelines for subject headings to be used at the Royal Library. Our aim is also to make these guidelines available and acceptable for all LIBRIS libraries.
We are very much concerned with the quality of the subject headings in LIBRIS. The major transformation that LIBRIS is to undergo has focused on the inconsistencies and poor quality of the subject headings used by us and by other LIBRIS libraries. Another problem is that very often the cataloguers chose to put their subject headings in local fields instead of using "general" fields in the LIBRISMARC format. Most of the headings in these local fields are exactly the same or made with only small variations. We need a great deal more co-ordination in this matter.
Another task which we find urgent is to make cataloguers use subject headings more often than today. Most libraries have been content to use only the classification codes in the records but we know now that subject headings, or subject headings together with classification codes, increase the possibility of the end-user to find what he or she is searching for. In order to make the SAB classification codes more useful as a search tool the LIBRIS Department has translated the codes into words and each registration of a code automatically generates the translation into a separate MARC field. For example, the SAB code for Philosophy is D - in LIBRIS you can search either on the classification code D or the word Philosophy.

Subject heading authority list

Swedish libraries have no common subject authority list for general use. Different libraries have their own lists and different ways of handling their indexing.
There is an index to the SAB classification system which we have been using as a subject heading list for the last few years. This SAB subject index has grown considerably over the last couple of years and has actually become an authority subject heading list and is now used not only by the Royal Library but also by other LIBRIS libraries.
The SAB subject index has classification codes linked to the subject headings, by no other hierarchy. This is a problem that we are aware of.

Conference on subject headings

BUS arranged a conference in February 1998 for the LIBRIS libraries about subject headings. The main goal with the conference was to make the libraries co-operate concerning subject headings. The conference was well attended with over 100 participants. The willingness to co-operate was clearly stated, but also the problems with different traditions and the lack of guidelines to follow.

Tasks ahead

Our main concern is to write guidelines that can be of help to the libraries in LIBRIS. First we want the guidelines to be accepted at The Royal Library and later at all LIBRIS libraries. Hopefully we will have them retrievable on Internet this autumn.
We are looking into other systems like Library of Congress subject headings for support. Multilingual subject headings is also an interesting matter which we will have a closer look into in order to find out if we can benefit from works done by other countries.
This Spring we visited BIBSYS in Norway and we are planning other library visits in the near future. Three of my fellow librarians and I went to Oslo to attend a conference arranged by Jon Anjer in May. It was a very successful and interesting conference with many speakers on different aspects of indexing. Another important task is to transform the subject heading list into a real authority list with a hierarchical structure.
We have also started to make weekly updating of new terms, terms suggested both by cataloguers at the Royal Library as well as by other libraries in LIBRIS. We think this is an important service to all kind of users to find relevant terms as quickly as possible.
See: www.kb.se/bus/aolista.htm Pia Leth,
Head of the Division of Bibliographic Development and Co-ordination (BUS) at the
Royal Library, National Library of Sweden

B. Library of the Riksdag

The preparation of the Swedish version of Eurovoc, that is, the thesaurus of the European Parliament and the Office for Official Publications of the European Communities, is completed. We await the printed version and will test Eurovoc in the indexing of parliamentary documents.

The work has been rather difficult, due to the complexity of the matter. Eurovoc is a multilingual thesaurus, developed originally in French. Necessarily some entity, some real object, must be the model. Different languages mean different ways of looking upon things, and they certainly do not always draw up the same conceptual boundaries. France with its law tradition, school system, etc. is not so easy to translate into Swedish equivalents. On top of this, Sweden is rather a new member of the European Communities and does not share the same history with the others. Many difficult concepts have not yet found a satisfying Swedish translation (it takes some time to develop a terminology), and many committees/organisations have never been given a Swedish name (only officials at the ministries needed them earlier and they got along with the English acronyms).

In the working out of the Swedish version - initially defined as to take charge of the scope notes and the non-descriptors, because the descriptors had already been translated - we learnt a hard lesson. It soon proved necessary to revise the translation of the descriptors, because of the very hard nature of the task. A thesaurus cannot be made in two steps: first the descriptors, then the rest. Those who take charge of the descriptors must also confront themselves with the scope notes, sometimes they definitely restrict the meaning of a concept, sometimes modify it, and the non-descriptors. You have to take everything into consideration: the concept in relation to its broader and narrower terms within the microthesaurus and the specific field, and furthermore the whole construction. Did not this term emerge somewhere else? A check shows: no, but a slightly different concept. Then you are confronted with a new problem. Have we used this term as a non-descriptor somewhere else? Yes, rightly so. Then where is the most appropriate microthesaurus for it? What about this concept? It did not appear in field number 12 and it is not in this field? Shall it be inserted as a non-descriptor because it is of importance from our perspective? How great freedom shall we take in creating non-descriptors? The questions are many, as are the decisions. If you got it wrong somewhere it will pop up somewhere else, as an omission or a repetition or still worse - because you will not understand it until it is too late - as overlapping, vague concepts. Ambiguousness is on the top of the list of what must not characterize a thesaurus, but it is hard to attain because of the complexity of the matter.

The question of automatic indexing arises now and then, indirectly from the work itself. The European Parliament conducted an investigation in 1995 but with poor results. At the Library of the Riksdag we have decided to enquire into the development of automatic indexing later this autumn to see if any progress has been made.

Elizabeth Lindkvist Michaeli
Library of the Riksdag, Sweden

Classification in the UK 1997/98

Much of the relevant information relating to classification in the UK has already been published in the Section's newsletter, circulated to members in May.

The proceedings of the 1997 "Dorking revisited" conference were published in the autumn of 1997 and copies were circulated to all participants and are still available for purchase from FID by others interested in obtaining a copy. All papers presented at the conference, for which disks providing revised copy were received by September 30th are included.

The Classification Research Group continues to meet on a regular basis at University College. It holds 5 meetings a year, and anyone who is interested may join the Group. The fee is �5.00 p.a. and all members receive the minutes of the meeting. About 50% of the membership comes from outside the UK so is not able to attend meetings on a regular basis. In the past year, the Group has discussed issues raised at the Conference, and has continued to act as a soundingboard for the revision of the Bliss Bibliographic Classification. The present state of publication with Bliss is that:

As far as DDC is concerned, for the UK the most important activity has been the development of the new Area Table for the re-organized local government regions of the UK.
The British Standards Institution is in the process of completing a "Pocket edition" of the UDC which will be on sale before the end of the year. This contains an introduction for novices and pared down version of the classification, together with an index, intended for personal use, the organization of computer files for personal computers, and for instruction in library schools. It will be of a size similar to a Penguin book, not exceeding 500pp. and will be priced at a competitive level.

BSI also produce an annual supplement to BS 1000M: the Medium edition of the UDC published in 1993. This comes out about March, and incorporates all previous editions together with the revisions in the current Extensions and corrections. The main volume together with this annual update will form the basis of a new edition of BS 1000M which should be published within the next two years.

Moving from the domestic to the international arena, the UDC Consortium has undergone a number of administrative changes in the past year, including the appointment of a new office manager at The Hague and a Research Assistant, based at UCL, to work on revision of the scheme. Negotiations are under way between BSI and Croatia on the production of a bilingual CD-ROM of the scheme. A Medium edition has been published in Portuguese and an abridged edition in Polish. Negotiations have begun for a new edition of the Guide and further translations, into Croatian and Portuguese, are in progress. It is possible that membership of the Consortium will be extended in the near future, as there has been considerable interest among a number of different parties in Russia to join, so there is hope that a Russian language member will join before too long.
There is a UDC Web site in the process of construction, and it should go live before the end of the year.

I.C. McIlwaine
University College London

Reports on classification from the USA

Activities related to classification research in the USA

  1. Metadata Development

Metadata is data about data. The term applies to any data used to aid the identification, description and location of networked electronic resources. The library community, which has always worked with metadata (i.e. cataloguing information) to aid in identifying, describing and locating its resources, was the first to approach the issue of the need for metadata standards if the potential of networked electronic resources was to be realized to its fullest.

Significant efforts have been made world-wide, with several being led by the US information professional organizations and associations.

OCLC has been leading in the development of the Dublin Core metadata standard. The Dublin Core is a 15-element metadata element set intended to facilitate discovery of electronic resources. Originally conceived for author-generated description of Web resources, it has also attracted the attention of formal resource description communities such as museums and libraries.
http://purl.org/metadata/dublin\_core\_elements

The Getty Information Institute has joined with The Research Libraries Group to launch the REACH (Record Export for Art and Cultural Heritage) project to address the museum community's need for an efficient and robust information network that facilitates access to information that museums wish to share with researchers and the public. The "REACH Element Set" is intended to provide a core set of elements that will be used in a database interoperability test-bed project.
http://www.ahip.getty.edu/index/warwick.html;
http://vads.ahds.ac.uk/metadataf5.html#5.1 Mapping to REACH Element

The Art Information Task Force (AITF), an initiative sponsored by the Getty Information Institute and the College Art Association (CAA), has developed "The Categories for the Description of Works of Art" which articulate an intellectual structure for the content of object and image descriptions.
http://www.gii.getty.edu/index/cdwa.html

The Visual Resources Association has developed "The Core Categories for Visual Resources" which is intended as a guideline for developing local databases and cataloguing records. The VRA Core element set contains two groupings of elements, the Work Description Categories (19 elements) and the Visual Document Description Categories (9 elements).

Encoded Archival description (EAD) is the emerging standard for archival finding aids which is supported by the Society of American Archivists and the Library of Congress. EAD makes it possible to provide access to archival finding aids in a platform-independent electronic format, using SGML (Standard Generalized Markup Language).
http://lcweb.loc.gov/rr/ead/eadhome.html

The LC Network Development and MARC Standards Office has developed "Dublin Core/MARC/GILS Crosswalk".
GILS Core Element to USMARC Mapping
(http://www.usgs.gov./gils/prof\_v2.html#annex\_b)

On the LC MARC website there are MARC mappings to other data standards, including Dublin Core, GILS, FGDC Content Standards for Digital Geospatial Metadata.
There is also a mapping from MARC to FGDC.

  1. Classification Research Related Workshops

ACM DL'98 (ACM Digital Library '98), Pittsburgh, June 27th Workshop on application of terminology and classification tools for digital collection development and network-based search (Over 30 participants are invited to attend the workshop).
http://www.alexandria.ucsb.edu/\~lhill/dl98\_workshop.html

The Workshop's Mission Statement is:

Goals of the workshop include:

  1. Move towards the initiatives toward interactive networked access to reference tools (terminology tools/services, concept tools/services and classification tools/services)
  2. A focus on current and planned developments to use terminology and classification tools in distributed information environments

The workshop resulted in two working groups to discuss further and work on:

  1. Registries for Knowledge Organization Schemes (KOS); Categories of KOS, Descriptions of KOS and Behaviours of types of KOS;
  2. Reference/Functional Model: High level object model (graphic representation), Interoperability with distributed searching, Support for distributed cataloguing, indexing and information systems.

After the workshop, the participants will work together and produce:

ASIS (American Society for Information Science) Special Interests Group/Classification Research Workshop, Nov. 1, 1997, Washington DC.
http://www.panix.com/\~schwartz/asis.sigcr/breakout.shtml

The Classification Research Workshop is designed to promote the exchange of ideas among active researchers with interests in classification creation, development, management, representation, display, comparison, compatibility, theory and application. Each year, researchers present their papers at the workshop. The 8th Workshop also devoted half a day to discuss topics of common interest, including: Classification Structures, Research Issues in Social Aspects of Classification, Challenges in Image Classification, Problems and Prospects in Thesaurus Construction and Exploring Bibliographic Classifications in New Environments.

Marcia Lei Zeng
Kent State University

Subject Indexing and Classification in the United States

SUBJECT INDEXING

Library of Congress Subject Headings

In 1998, the Library of Congress celebrates the centennial of Library of Congress Subject Headings (LCSH). On June 27th 1998, during the American Library Association, a ceremony commemorating the occasion was held in the Great Hall of the Library. In the summer of 1898, a new dictionary catalogue was begun at the Library of Congress, with the integration of alphabetically arranged author, title and subject entries. LC cataloguers began to select and apply subject headings from copies of the 1895 A.L.A. List of Subject Headings and added new ones as they were needed in cataloguing the Library's collection. This marked the beginning of LCSH. Over the past century, twenty editions of the list have been published. Concurrent with the celebration, the 21st edition, a five-volume set, was designated the centennial edition of LCSH. It was published in June 1998.

A major change in LCSH is the implementation of form headings and form subdivisions. Previously, form headings/subdivisions were indistinguishable from topical headings in codings in the USMARC formats. The separation improves the faceting principles of LCSH. The electronic version of LCSH now contains not only records for main headings and many heading/subdivision combinations, but also separate records for free-floating subdivisions. Records for approximately 3,000 topical, form and chronological free-floating subdivisions are now being created and incorporated into the main file.

Medical Subject Headings and Metathesaurus

Medical Subject Headings (MeSH) published annually, continues to serve as an indexing and cataloguing tool for medical literature. A major change took place recently. The geographic and form subdivisions have now been taken out of the subject heading strings and assigned in separate fields. This change signifies a move of MeSH towards a more faceted, postcoordinate system. ,p> The Metathesaurus, a database of information on concepts that appear in one or more of a number of different controlled vocabularies and classifications used in the field of biomedicine, continues to expand. The 1998 Metathesaurus contains 476,313 concepts and 1,051,901 different concept names from more than 40 vocabularies and classifications.

Sears Subject Headings

Sears Subject Headings continues to serve as the subject access vocabulary for smaller libraries. The 16th edition was published in 1997.

CLASSIFICATION

Library of Congress Classification

The LC classification continues to serve academic and research libraries in the United States in organizing their collections. The most important recent development is the conversion of the schedules and tables into electronic form, using the USMARC Format for Classification Data. This machine-readable form is used within the Library for processing its materials. It is also available outside in the from of a CD-ROM product called "Classification Plus". This product contains both the LC Classification and the LC Subject Headings, with links between many class numbers and their corresponding subject headings. New editions are now produced from the electronic version at more regular intervals than before.

The LC classification has been in use by the Library of Congress and other libraries for a century. Nevertheless, until 1997 it had been an incomplete scheme because many of the law schedules had not been developed. Now, with the publication of the schedule for Subclass KZ (Law of Nations) the Library of Congress Classification is now complete, except for subclass KB which is to be published in 2000.

National Library of Medicine Classification

The most recent edition, the 5th, was published in 1994.

SUBJECT DATA IN METADATA RECORDS

To study the application of subject data, including classification and free-text and controlled vocabularies in metadata records, particularly the Dublin Core records, the Subject Analysis Committee of the American Library Association has appointed two subcommittees with the charge of recommending the most effective methods of supplying subject data for the representation of the contents of World Wide Web resources.

Lois Mai Chan
University of Kentucky. USA

Other news

At the IFLA committee meeting Ala Sabina Mieziniene, Vilnius University, from Lithuania, was welcomed as an observer, and she reported on activities related to classification and indexing in her country. She stated that an abridged edition of UDC was in use there, but that the National Library used had produced full editions of Classes 1, 2 and 8 and parts of class 3. The subject headings list used nationwide was that produced in the National Library.

Bill Lamble, an observer of the Committee reported from Australia that the Library and Information Service of Western Australia has recently started a project of automating inhouse card indexes and incorporating the information in a bibliographical database, and has contributed the following report:

Finding and Identifying Index Records in Innopac

Finding and Identifying Index Records in Innopac

New inclusions in LISWA's Catalogue

Since 4 May 1998 clients searching Innopac will have noticed that retrieved sets of records include the normal records for bibliographical items (books, serials, sound recordings etc) as well as index records for things like periodical articles, book reviews, poems, songs, maps and illustrations that are contained within other publications.

This is an exciting development for clients who previously needed to visit the Alexander Library Building to find index records in one of the many different card indexes round the building. Now all the indexes will be available on line and in one sequence - interspersed with the other bibliographical records.

These Index records are very easily distinguished from normal records for items.
Each displays with a highlighted message labelled "Found In" where the details are cited for the publication that contains the indexed information. Clients wishing to see the indexed information can apply for an inter-library loan if the source cited is a monograph or may request a photocopy if the source publication is a periodical or newspaper.

Types of Index Record

The database now includes index entries for newspaper articles on a wide range of topics - but chiefly of Western Australian interest. General index entries may also contain facts which are very frequently asked for in libraries. Facts such as the Seven Wonders of the Ancient World, or the latest winner of the Nobel Prize for Literature are provided. These records can easily be distinguished from normal entries for monographs or serials by the word "fact" which appears in parentheses at the end of the title e.g. [Seven wonders of the ancient world] [fact] .

Viewing different parts of the database

Clients wishing to access records for Oral History alone, or for Maps, or Australian Poetry may elect to search just that part of the database. From the first menu in Search clients may choose to Change Library Catalogue. The following menu appears: 1 > Entire Collection
2 > Alexander Library Building
3 > Public Library Stock
4 > Private Archives Collection
5 > Map Collections
6 > Pictorial Collection
7 > Oral History Collection
8 > Song/Instrumental Index
9 > Australian Poetry Index
10 > Ship Index

This menu enables clients to choose to search any one of these special collections or indexes or they may decide to look for index entries in the Entire Collection and Alexander Library Building.

More about each of these special indexes will appear in future issues of Knowit.

Information received

From Germany, we have been informed that Bernd W.J. Lorenz has prepared a nine page pamphlet: The Regensburg Classification Scheme, a short survey. It will be useful for students and persons who are entirely unfamiliar with the Regensburg classification. It is available from Koordinierungsstelle für den Klassifikationsverbund, Universitätsbibliothek Regensburg, D-93042 Regensburg, Germany.

In Russia, a conference on "Actual problems of subject indexing and subject access systems" was held at the National Library of Russia, April 27-29, 1998. It was sponsored by the Russian Library Association and the National Library of Russia. Approximately 100 people attended, with most from Russia.

The conference was held in celebration of the 30th anniversary of the Research Group on Subject Analysis and Subject Cataloguing. The Russian MARC formats for authority records, both for internal use at the NLR and for communications were presented, as was a translation of Guidelines for Subject Authority and Reference Entries.

A fuller report of the Conference is to be published in International Cataloguing and Bibliographic Control. The papers are to be available, in Russian and English, on a Web page devoted to the conference:
www.nlr.org

Report from the ICNBS (International Conference on National Bibliographic Services) Copenhagen 25-27 November 1998

This conference, the 2nd International Conference on National Bibliographic Services was auspiced by IFLA and three Danish library institutions.
The first conference which was held in Paris in 1977 (at Unesco) resulted in very important recommendations on National Bibliographies

Now was the time to look into these more than 20-year-old recommendations and see in what way they needed updating.

You can find all about the conference on IFLA's homepage
http://www.ifla.org/VI/3/icnbs/beam.htm

The conference was very well organized and all participants generously taken care of.
The participants got useful background papers including the Unesco 1977 recommendations, beforehand. The pre-conference booklet also included the paper on subject retrieval in national bibliographies by I.C McIlwaine and Lois Mai Chan, which has been circulated to the members of our section. There were papers on the inclusion of information covering electronic resources in national bibliographies, national bibliographic agencies and the book trade, bibliographic control activities in Southeast Asia, among others.

At the conference there were several very interesting lectures. For example how do you exclude material from the national bibliographies when you most of all want to include everything at the same time as the amount of publications is growing?
The question of how to handle on-line documents was also raised.

Legal deposit plays an important role and it is vital that co-operation should be of benefit for the publishers as well as the libraries. Therefore the national library agencies must be timely as well as producing bibliographic descriptions of high quality in order to be of use to the publisher and to get publishers to co-operate

The most important issue, of course, was to study the recommendations of 1977 in order to update them.
Therefore participants were divided into discussion groups with certain topics to discuss:

How far should bodies other than the national library be involved in the production of the national bibliography?
Is the role of the national bibliography being overtaken by the Internet?
What is the future of standards in the Internet age?
Can national bibliographic agencies afford to record everything that is published?
In what way can national bibliographic agencies in the richer countries assist those in the less rich countries
What is the future of legal deposit?
What is the future of the printed national bibliography?

So what recommendations did the conference agree upon?
A draft was discussed on the last day of the conference which stressed the importance of the national bibliography as a major instrument in ensuring a full record of the national published heritage and implementng full bibliographic control.

The draft covered recommendations on:
Legal deposit
Coverage of the national bibliography
The presentation and timeliness of the national bibliography
Standards
Future activities

Pia Leth

Subject indexing and classification developments in France, 1997-1998.

Concerning Classification, the major event of this year is of course the translation of the 21st edition of the Dewey Decimal Classification. First, with the publication of an abridged edition (editions Cercle de la librairie), and especially with the launching of the complete edition in French, the first since 25 years, which will take place officially during this IFLA Conference. As the result of a partnership between Canada (ASTED, NLC, BNQ) and France (BnF), this co-edition is a significant experiment which will be reported on during the second workshop of our section (Thursday 20.08, afternoon). It gives us a global vision of the DDC, thus enabling us to better assess what seems to us the principal problems and to formulate some suggestions.

From the national point of view, the DDC is currently widely used to establish call numbers and to localize documents, but often with many adaptations, indispensable for the most part, and not like a "standard". On the other hand, traditionally like any classification in France, it plays only a very small part in subject access. For the moment in the BnF, only the collections on open access are indexed with Dewey and the numbers thus provided, which are translated into words to facilitate search, form the basis of the work of setting up the Dewey BnF authority file. The remarks occurring to us during the second reading work enriched our reflection on its future use, which will perhaps tend towards less precise Dewey numbers, as there is a close connection with the national subject heading system RAMEAU.

Lastly, within the Permanent UNIMARC Committee (PUC), the institution also participates in the development of the UNIMARC Format for classification data. Still on the subject of formats, the BnF developed in 1998 in connection with external libraries a French interchange format for authority records in UNIMARC. The year 1997-1998 will have been, moreover, particularly important for formats and systems. Concerning university libraries, the project of Système universitaire de documentation (SU, University System of documentation) has been undertaken since the beginning of 1997 with the Agènce bibliographique de l'enseignement supérieur (ABES, Bibliographical Agency of higher education). It will be operational in 2000, and will integrate in UNIMARC all the catalogues of existing networks and will become the unique system of all French university libraries.

At the BnF, the data of all existing catalogues and bases have been converted into the new INTERMARC Integrated format, which gives notably a new structure to the BnF authority records, and have migrated towards the new Système d'information (SI, Information system). With more than 10 million on-line records, the SI will be operational for consultation by the time of the opening of the rez-de-jardin research reading rooms in October. It should be operational in production at the end of 1999.

In SI as in SU, the standard of subject access is RAMEAU, the national subject heading language created in 1980 and developed on the network since 1987. From 1999 and in conjunction with the new format of the SI, its management in the BnF will be better focused on proper "subject" headings (common nouns) since proper names (authors, corporate bodies, titles), handled by the responsible authority files, will have a single form of authority and will be usable as a subject as well as author point of access. RAMEAU which is used by all types of libraries, for all types of documents, is presented and described in the IFLA document on "Principles underlying SHLs". The part relating to it was reviewed in its entirety and updated at the end 1997-beginning 1998 within the Working Group of our section. New instruments of work and management relating to our SH Language are now under way:

By the beginning of 1998, a new RAMEAU tool had already been developed on CD-ROM of the BnF authorities, with an English-French index allowing direct search by LC headings equivalent to the RAMEAU ones, and a word index allowing search by words of these English subject headings.

These concordances and the experience gained since the beginning of the nineties were particularly useful for the "multilingual European thesaurus" project which will be presented during the Open forum of our Division (Monday 17.08 morning, Magda Heiner-Freiling). Much work has been undertaken in this since the end of 1997 within a CoBRA+ Working Group (BL, BnF, BNS, DB), this co-operation between European national libraries aims to exploit the richness and the flexibility of the national subject heading languages in order to bring about multilingual subject access to the bibliographical databases by the establishment of links of equivalence between the headings of the subject heading systems used.

Lastly, and still on the topic of the co-operation, it should be noted that RAMEAU also pursues active co-operation with other countries in Europe (Polish libraries and the thesaurus KABA, public libraries of the French Community of Belgium) and with other French-speaking countries (intended renewal at the end of 1998 of the Convention with the Bibliothèque de l'Université Laval, Répertoire de vedettes-matière).

Max Naudi
Bibliothèque nationale de France