When it comes to creating a library automation infrastructure in 2010, there are many models in play. Should libraries continue to rely on integrated library systems installed locally in libraries or consortia? Will some shift to relying on a globally distributed infrastructure through OCLC’s Web-scale Management Services? Might some libraries rely on vendor-hosted arrangements through software-as-a-service?
The same kinds of questions arise in the way that libraries deal with the bibliographic records that describe their collections. The longstanding copy cataloging model involves libraries relying on external bibliographic sources for items that have previously been described, bringing copies of records into their local system. Only when no acceptable copy of the record can be identified will a library perform original cataloging, creating a new record from scratch. OCLC member libraries can rely on the massive WorldCat database of bibliographic records to achieve a very high ratio of copy cataloging relative to original cataloging. Outside the OCLC fold, libraries depend on other bibliographic resources such as the catalogs of their national libraries, or a more peer-to-peer model of deriving records, often based on informal terms, from other libraries that happen to have servers supporting the Z39.50 protocol. In this month’s newsletter, we discuss BookWhere, a software tool that helps many libraries to obtain the records they need to describe their collections.
Today, several different trends are underway with major implications for how the library of the future will deal with metadata. While I don’t claim any specific insight into the nuances of cataloging, it does seem to me that new models of automation will impact cataloging workflows and that new development in the cataloging and metadata arena may require fundamental changes in library automation systems.
The metadata model inherent to OCLC Web-scale Management Services radically simplifies the workflow involved in copy cataloging. Traditional ILS models involve downloading MARC records from external resources and importing them into a local system, often making changes to the records according to local cataloging practices. This approach offers the benefit of customized records for a given library, but at the cost of each library making changes independently of others that use the same record. WMS eliminates the need for local copies of records, but means that all libraries must accept a given record or enhance it for the benefit of all.
As new generations of library automation software become more oriented toward Web services and other more current technologies, handling bibliographic records in MARC communications format rather than XML becomes more problematic.
While it’s possible, and routinely implemented, to translate MARC records into MARCXML, many idiosyncratic characteristics remain. The record structures associated with MARC communications format, such as BER (Basic Encoding Rules) and ASN.1 (Abstract Syntax Notation), which worked well for storing records on magnetic tape and for producing catalog cards, don’t always fit well within technology platforms based on XML structures which are moving increasingly toward semantic Web concepts of linked data.
Resource Description and Access (RDA) proposes to modernize some of the cataloging rules used for creating bibliographic records. Ideally implemented in an XML document structure, these rules have been applied to bibliographic records in MARC formats, with mixed results. While many perceive the advantages in a more updated set of cataloging practices, the immense installed base of bibliographic records mean a very difficult transition. The current generation of integrated library systems, designed around bibliographic records coded according to AACR2 can be adjusted to accept records created under the new RDA conventions, but cannot necessarily take full advantages of the benefits envisioned without a more wholesale conversion.
These are just some very basic examples of the interrelations between the ways that libraries deal with metadata and the technology systems involved in describing and providing access to library collections. New conceptual models are unfolding in both the metadata and in the library management systems arenas.
Of course the major players in both sides are well aware of each other’s issues and concerns. Hopefully we’ll eventually end up with technology products that take full advantage of new metadata practices and that metadata practices will continue to evolve in ways that can be best exploited with technology platforms to provide the best value to library users.