The aim of this project is to assist researchers to locate, in UK university libraries, works written in Arabic and Persian. At present tracing such material can be difficult, particularly as much of it is still only represented in local manual catalogues.
Under this project, the participants will add records for around 60,000 Arabic and Persian items to local, national and international catalogues. The records will be created on RLIN, the bibliographic database of the U.S.-based Research Libraries Group, with the bibliographic description entered in both romanised (transliterated) script and in the original Arabic script. Not all libraries are currently able to make use of the Arabic script in their local system, but the aim is to ‘future proof’ the conversion work at a time of rapid change in the area of computer handling of non-Roman script.
While the RSLP funding was a not-to-be-missed opportunity, 1999 was an ‘interesting’ time to be planning a retrospective conversion programme for Arabic script material. Unicode wasn’t yet widely available and none of the main library management systems could accommodate Arabic script. However, developments in this area were imminent and it seemed undesirable to create computer catalogue records solely in transliteration (that is, with the original script transcribed into its romanised equivalent). To ‘future proof’ the conversion work, the libraries needed to be able to create, edit and store records containing both the usual transliterated elements and a parallel set of fields holding the bibliographic description in the original script. Fortunately this was possible using RLIN.
Most UK libraries use RLIN for cataloguing by simply importing copies of matched records into their local system. It is possible, however, to use the RLIN interface to create, derive and amend catalogue records within the RLIN database. This facility offered the partner libraries the opportunity to create and edit records with both romanised and Arabic script fields, with the records immediately available for use by the other libraries.
Those partner libraries that use the MARC21 bibliographic format can import into their local database the full records they create, even if they cannot as yet edit or display the script fields locally. Libraries that are using the UK MARC format have the problem that the MARC21 to UK MARC conversion programs cannot handle the script fields. These libraries can only load the transliterated part of the records into their local catalogues now, and have to defer taking the full records either until they move to MARC21, or acquire the facility to locally edit the original script in the records.
The technical developments believed three years ago to be on the horizon are indeed materialising. Two of the partner libraries, Durham and SOAS, have recently implemented the Innopac Arabic script module, and so are able to display records in the original script on their local OPACs.
Earlier this year, RLG released a new version of the RLIN end-user interface, Eureka, which has the ability to display records in original script. This means that the Arabic script records for the material catalogued under the project can be displayed in Eureka.
Work on the project started in February 2000 at Manchester, with the other partners starting at intervals over the following year. Work will continue until October 2002.
Hit rates against the RLIN database have been good - averaging over 80%. However, only around a third of the matched records have been found to include Arabic script, so project participants have had to add original script to the remaining two-thirds of matched records as well as to the original cataloguing.
As of February 2002, records for over 38,000 Arabic and Persian items had been added to local catalogues.
Project manager
Shirley PerryTel: 0161 275 8715
E-mail: shirley.perry@man.ac.uk
http://rylibweb.man.ac.uk/rslp_mes/
Content: Gill Davenport
Last updated 3 July 2002