The Semantic Web - Social Semantics: The Search for Meaning on the Web - page 56

lieutenant of the Cyc project, the last-ditch Manhattan project of classical artificial
intelligence (Guha and Lenat 1993). There are nonetheless some key differences
between semantic networks and RDF, as RDF was built in accordance with the
Principles of Web Architecture that were explained in Chap. 2, as detailed in the
next subsections.

3.2.1 RDF and the Principle of Universality

Semantic networks fell out of favor because of their use of ambiguous natural
language terms to identify their nodes and arcs, which became a problem when
semantic networks were transported between domains and different users, a problem
that would be fatal in the decentralized and multi-lingual environment of the Web
(Woods 1975). According to the Principle of Universality, since a resource can be
anything , then a component of the knowledge representation language should be
considered a resource, and thus can be given a URI. Instead of labelling the arcs and
nodes with natural language terms, in RDF all the arcs and nodes can be labelled
with URIs. Although few applications had ever taken advantage of the fact before
RDF, URIs could be minted for things like the Eiffel Tower qua Eiffel-Tower, an
absolute necessity for knowledge representation. Since the sense of statements in
knowledge representation is usually about content in the world outside the Web,
this means that the Semantic Web crucially depends on the rather strange fact that
URIs can refer to things outside the Web.
This does not restrict the knowledge-representation language to merely referring
to things that we would normally consider outside of the Web, since normal web-
pages use URIs as well, and so the Semantic Web can easily be used to refer to
normal web-pages. This has some advantages, as it allows RDF to be used to model
the relationships between web-accessible resources and even mix certain kinds of
relationships. This sort of “meta-data” is exemplified by the relationship between a
web-page and its human author, in which both the author and the page would both
be denoted by URIs using RDF. Lastly, this ability to describe everything with URIs
leads to some unusual features, for RDF can then model its own language constructs
using URIs, and make statements about its own core language constructs. However,
just as all components of RDF may be considered resources, just as all resources
may not have URIs, all components of RDF may not have URIs. For example, a
string of text or a number may be a component of RDF, and these are called literals
by RDF. In RDF specified anonymous resources are not given a URI, and these are
called blank nodes . Yet it would be premature to declare that the deployment of
URIs in RDF signals a major improvement over natural language labels, for URIs
can be just as ambiguous as natural language labels by themselves. However, various
theories of semantics as well as engineering like the 'follow-your-nose' principle
were theorized to solve the problem of ambiguity.