Page 132 - Building Digital Libraries

Metadata Formats

replace AACR2, but to eventually become the dominant set of practices for
describing all bibliographic metadata. RDA attempted to shed the MARC-
isms, and focus on modeling questions to allow the concepts to be applied
outside of traditional MARC, but into other library schemas (like EAD,
TEI, MODS), but outside of the library community to support archive and
museum cataloging, as well as the emerging digital library space. Addition-
ally, it was hoped that RDA would provide the foundation upon which a
new bibliographic model could be developed: one that went beyond MARC
and that would usher libraries into the world of semantic web description.
And while one could certainly argue whether these goals were achieved,
RDA has successfully supplanted AACR2 as the primary data model for
describing MARC data, and it has had a tremendous impact on the Library
of Congress’s semantic web efforts, including the continued development
of BIBFRAME.
A better question to ask is whether RDA is still relevant within today’s
digital environment—particularly given the fact that AACR2 was created
for a print-centric world where the only catalogs contained printed cards.
In fact, Tennant seems to make this argument in a later article where he
4
steps back from calling for the death of MARC, and argues for a new way
of thinking about bibliographic infrastructure. Today organizations like
the Library of Congress question the need for traditional controlled access
points and the future of metadata creation, leaving libraries to struggle
6
5
with how digital materials should be described and be accessed within local
and remote systems. This does not mean that much cannot be taken from
the current descriptive rules and utilized as an organization begins its digital
repository efforts, but an organization must recognize that using only the
current bibliographic descriptive rules will be insufficient. However, before
taking a closer look at some of the common XML metadata formats that are
supported using today’s digital repository software, a very brief look at the
MARC metadata schema is in order.
Finally, one of the significant changes in digital repository develop-
ment over the past ten years has been the increasing support for semantic
data, and the ability to model bibliographic data utilizing RDF (Resource
Description Framework). In the early and mid-2000s, libraries primarily
selected a single metadata framework, and utilizing that framework, all
content would need to be captured. Admittedly, some tools like DSpace
provided the ability to create custom fields for local use, but ultimately, at
the system level, all data was mapped into a single metadata format. Today,
this is no longer true. Systems like Fedora 4 and Ex Libris’s Rosetta software
have been developed using linked data and semantic web concepts as their
foundations. This means that data is no longer limited to a single metadata
schema or data model; instead, through the use of multiple schemas and
data namespaces, materials can utilize multiple metadata schemas through
RDF. In this section, I will detail a handful of the common metadata primi-
tives that are still utilized within digital repositories, but I will also discuss
the emerging presence of linked and RDF data, and how this is changing

117