Page 156 - Building Digital Libraries

Metadata Formats

community see the implementation of systems utilizing a BIBFRAME model
as simply a matter of time. While the MARC standard is still being main-
tained and grown, the bulk of the development work within the standards
community appears to be focused on BIBFRAME and RDA. At the same
time, libraries have a significant amount of legacy data and infrastructures,
resources that simply cannot be left behind. And in this, there is the chal-
lenge. While BIBFRAME development continues within the community,
practical development and implementation have yet to move outside of the
research and development phase.
While BIBFRAME may never replace the MARC format, it has been
hugely successful in underlining the importance of linked data to the future
of libraries. BIBFRAME was developed as a semantic language, and the need
for linked data infrastructure and URIs has highlighted a glaring omission
in the library community’s current infrastructure. This has led to the rapid
development of semantic endpoints by organizations opening up their con-
trolled vocabularies for machine-readable processing. Organizations like
the Library of Congress, the Japanese Diet Library, OCLC, and the Getty
Research Institute have made tremendous strides over the past few years
in developing an infrastructure to support the rich linking of string data to
semantic resources.
As this infrastructure has developed, the library metadata community
has moved to enable linked data support within legacy data formats like
MARC in order to help ease the transition to a new data model, and hope-
fully spur innovation within current library systems. Tools like MarcEdit,
Catmandu, and OpenRefine have developed special tool sets to support the
reconciliation of library data with these new linked-data services.
As of this writing, the jury on BIBFRAME’s success as a format is still
out. Will it become a predominant data format like MARC? Will it replace
MARC? BIBFRAME’s primary success and greatest long-term benefit to the
library community will be as an experimental format that pushed semantic
and linked-data concepts into the forefront of library metadata discussion.
Through these discussions, libraries are now taking the development of
semantic web infrastructure seriously, and are actively looking at how they
can support the enhancement of legacy data to enable future innovation.
In the long term, I think that these successes will be the lasting legacy of
the BIBFRAME development, and may ultimately be what spurs libraries
to ensure that their digital repository systems natively support the semantic
web, allowing for their repositories’ interoperability with systems and tools
well beyond the traditional library community.

Domain-Specific Metadata Formats
Unfortunately, this chapter just scratches the surface when it comes to the
metadata formats that are currently in use within the digital library com-
munity. While Dublin Core, MODs, and METS are likely the most widely
used today, these represent general-purpose formats. Within the various

141