Page 175 - Building Digital Libraries

CHAPTER 7

or are actively building the infrastructure necessary to promote linked
data use and reuse in a production environment. This work is hard and
expensive, and this is one of the reasons why linked data remains an open
opportunity rather than a realized promise. Even VIAF, a project that OCLC
has been building and maintaining for years, lacks some of the necessary
infrastructure to be a part of the current linked-data infrastructure. As we
look to the future, how digital libraries support the creation of linked data
services and take advantage of them will be interesting to follow.

Sharing Metadata
Fortunately, nearly all digital repository platforms provide some method of
sharing metadata. Generally, these options are limited to harvesting proto-
cols, enabling users to download subsets of an organization’s metadata for
use within other systems. While this information isn’t generally available in
real time, this process has fueled a large number of discovery solutions that
have been developed around digital libraries. More recently, libraries have
been embracing standard HTTP-based operations, and developing REST-
based services that utilize JSON as the primary method of communication.
Moreover, new HTTP-based protocols like ResourceSync and SPARQL
5
6
likely represent the next generation of harvest and query operations within
the digital library space. These services are unique in that they seek to
leverage established HTTP patterns and norms, rather than creating library-
specific communication standards, thus making content more accessible
to those outside of the library domain. However, since current-generation
digital repository platforms primarily support OAI harvesting as a meth-
odology for metadata distribution, much of the remainder of this chapter
will focus on the tools, methods, and protocols necessary to crosswalk har-
vested content to different metadata schemas; on issues related to metadata
crosswalking; and on the protocols used to support metadata harvesting on
current digital repository platforms.

XSLT
When considering XML metadata crosswalking, very likely one will be
working with XSLT (eXtensible Stylesheet Transformation). XSLT is a W3C
technology designed to work with XSL (eXtensible Stylesheet Language),
which is essentially a stylesheet language for XML. XSLT was originally
designed and is used primarily for the transformation of an XML document
from one document language to another. On the Web, XSLT is utilized to
transform XML documents into HTML documents for public display.
XSLT itself defines a list of vocabularies that can be used to manipu-
late an XML document. By itself, an XSLT stylesheet is no more than an
XML document. Transformation commands are constructed as tags within
the document, utilizing XPath to extract and reformat individual data
160