|Cataloguing and Retrieval- Local History Digitisation Manual|
Existing cataloguing standards
Recording technical/copyright information
To ensure ongoing use of the digital images created during a digital project, it is imperative that rich, reliable catalogue information is available and linked to the digital image. However in large collections it may be too costly and time consuming to create new catalogue records for each item. In some cases existing records will be held in a proprietary library catalogue software system and it may be possible to either use this system or to export the information into another system to provide search and retrieval of the digital files. The most important decision will be what type of catalogue information to keep for each digital item and how this will be stored and accessed.
The IT industry uses the term 'metadata' to describe data about digital data, but the concept is the same as library catalogue information, the only difference being the manner in which it is stored. The idea of metadata arose from the perceived need of the IT industry for an agreed set of catalogue fields held in a standard format which could be understood and accessed by any computer on the Internet. If these standards were widely implemented, then theoretically any information made available over networks could be searchable by any computer which recognised the way the data was structured and could therefore identify the item in question. Metadata allows for the possibility of worldwide searching for digital files across the Internet or at least within specific communities.
Adherence to either community metadata standards or, at a minimum, internally consistent data standards remains the surest guarantee of future data viability. (Kenny & Rieger, 123)
It is therefore imperative that catalogue or metadata standards which are appropriate to the current, and if possible future, needs of the users are established and adhered to for digital items. Digital files also create additional metadata requirements as technical information about the files themselves should also be stored to assist with future use and migration of the data.
Unfortunately there is "no
single accepted standard for cataloguing all material in electronic form". (Lee,
106) There are a number of different metadata standards, but one which has been
widely accepted as a valuable basis for wide interoperability is the 'Dublin
Core' standard. Dublin Core provides a set of simple descriptive elements for
the resource discovery of document-like images items on web. The elements were
meant to be simple enough to allow non-expert authors to create their own
metadata. The elements can be embedded in HTML code and if widely used would
allow for more effective resource discovery across the web. The 15 Dublin Core
metadata elements are:
For further information and descriptions of the contents of each Dublin Core metadata element see http://dublincore.org/documents/dces
The Dublin Core elements have been adopted and adapted for use by a number of different industries and there are now tools which can automatically create Dublin Core records. See http://dublincore.org/tools
Adoption of an existing metadata standard is likely to assist with future interoperability with other collections, but it is important that the standard chosen meets the specific needs of the collection. The most important consideration is deciding what information needs to be recorded to meet the needs of cataloguers, users, technical experts and administrators. (Lee 108)
All libraries will be familiar with MARC (machine-readable cataloguing) standards, and most Victorian public libraries catalogue their local history collections into their main catalogues in MARC format. MARC data can be mapped to Dublin Core metadata, and this approach underpins the National Library of Australia's Picture Australia database.
The State Library of Victoria enters digital images into its catalogue with full MARC AACR2 catalogue entries. It currently has around 175,000 images available through both the State Library of Victoria catalogue and Picture Australia.
When creating new records, there are some economies of scale which can be achieved. Many items are quite similar and have characteristics in common. A cloned record will build on information assembled for another, similar, item.
The use of standard thesauruses when creating catalogue records will speed up the process of cataloguing and ensure standardised data and therefore more efficient searching. It will also assist with any future interoperabilitily between collections using the same thesauruses. The Australian Pictorial Thesaurus (APT) is an online thesaurus of Australian subject terms for searching and indexing pictorial and other original materials collections in Australian libraries and museums. The APT has been developed by the Council of Australian State Libraries (CASL) to provide a new image indexing standard. It may be appropriate for public library local history collections to adopt APT for their digital image collections. For more information see: http://www.picturethesaurus.gov.au
As digital files are not readable without technical intervention, the availability of technical information about the files will be imperative for future management of the collection. It is therefore important to record relevant technical data about digital files at the time they are created when it is readily available. This will allow for future identification and management of specific items and will assist with long term management and data migration for the whole collection.
Consideration should be give to including information such as; date of creation, file format, image resolution, bit-depth, URL, creation device, file size, size of original plus other technical information where appropriate. Standard vocabularies or thesauruses will also assist with ensuring consistent use of terminology. Well structured, high quality technical information may enable automatic data migration or manipulation in the future and in a large collection this is likely to be invaluable.
The inclusion of information on the copyright status of the items will also assist future use and management of the collection. If agreements with specific usage terms have been entered into, these should be recorded in the metadata associated with the item to ensure there is no future breach of copyright in relation to the item. If a copyright owner has been located this and other pertinent information may also need to be included. The type of copyright data to be kept will be depending on the current and future management needs of the organisation and any agreements entered into with copyright owners.
If smaller digital collections hold their data in a format which can be accessed through other collection interfaces it is likely to reach a wider audience. The use of metadata standards such as Dublin Core and thesauruses such as APT make it more likely that the collection will be interoperable with other collections. The Picture Australia service is an excellent example of an interoperable system. Picture Australia uses metadata harvesting software to collect data held in a standard form from participating collections. The digital images remained stored by the organisation who manages the collection, but users can come to a central search interface and access images from a large number of collections. Planning for interoperability through the use of standard metadata and image file formats is the most effective way of ensuring the widest possible audience for the collection and the ability to migrate data in the future if necessary.
Some Victorian public
libraries are currently working on providing images to Picture Australia. To do
this it was necessary to ensure that existing catalogue data could be mapped to
a format appropriate for use by Picture Australia. Catalogue information which
has been tested for inclusion was held in databases using Microsoft Access
database and 'In Magic'. For further information on progress with these
projects, contact Vision Australia at email@example.com or
Maribyrnong Library Service at firstname.lastname@example.org