Page 161 - Building Digital Libraries

CHAPTER 6

may relate to it. In reality, automated advertisement-blocking software is pri-
marily making its best guess regarding the nature of the examined content.
For catalogers, the problem is the same. Automated bibliographic creation/
editing processes still require a great deal of human intervention to create
the necessary rules for a machine to interpret a metadata set. Without these
rules, a machine could certainly parse a metadata set, but it would have no
way of building or interpreting the relationships of this bibliographic data
to other metadata content.
The semantic web, then, is about connecting these dots to give machine
processes the ability to determine and infer relationships and meaning
about the data that it describes. If this sounds a lot like metadata for one’s
metadata, you would be essentially right. RDF, the Resource Description
Framework, is one of the cornerstones of the semantic web initiative. RDF,
and similar serializations, provide the Web with a common descriptive
language. In theory, RDF provides a common descriptive framework that
can be used to “wrap” an existing set of metadata/data to provide the miss-
ing information needed to give machine processes the ability to draw rela-
tionships between heterogeneous datasets. RDF, coupled with OWL (Web
Ontology Language), make up two major components of the semantic web
framework that is designed to bring out the relationships between data on
the Web.
So, what does this look like? Well, it depends on the serialization and
the namespaces being used. The Dublin Core community has a long his-
tory of utilizing RDF as part of its metadata description language. As early
as 2007, the Dublin Core Metadata Initiative produced a draft document
entitled “Expressing Dublin Core Metadata Using the Resource Description
Framework (RDF).” This was its most detailed document to date discussing
19
how Dublin Core metadata could be expressed within the RDF metadata
framework. The document illustrates how Dublin Core can be encoded to
ensure that nonhuman processes can understand the important concepts
related to the metadata record—for example, details relating to classifica-
tion, ontologies, and relationships (both hierarchical and relational) to
other items. At its simplest, the RDF encoding of Dublin Core data can look
something like the following:

<?xml version=“1.0” encoding=“utf-8”?>
<rdf:RDF xmlns:rdf=“http://www.w3.org/1999/02/22-rdf-syntax-ns#” xmlns:dc=“http://purl.org/dc/
elements/1.1/”>
<rdf:Description>
<dc:title>Report of the exploring expedition to the Rocky Mountains in the year 1842 : and to
Oregon and north California in the years 1843–44 /</dc:title>
<dc:creator rdf:resource=“http://authorities.loc.gov/” />Frémont, John Charles,1813–1890.</
dc:creator>
<dc:creator rdf:resource=“http://authorities.loc.gov/” />Torrey, John,1796–1873.</dc:creator>
<dc:creator rdf:resource=“http://authorities.loc.gov/” />Hall, James,1811–1898.</dc:creator>
<dc:creator rdf:resource=“http://authorities.loc.gov/” />United States.Army.Corps of Engineers.</
dc:creator>
<dc:type>text</dc:type>
<dc:publisher>Washington : Blair and Rives, Printers,</dc:publisher>

146