The Semantic Web - Social Semantics: The Search for Meaning on the Web - page 55

Ultimately, the project of artificial intelligence to design a single knowledge
representation system suitable for creating human-level intelligence has not yet
succeeded and progress, despite occassional bursts of enthusiasm, is doubtful at
best. With no unifying framework, the field of artificial intelligence itself fragmented
into many different diverse communities, each with its own family of languages
and techniques. Researchers into natural language embraced statistical techniques
and went back to practical language processing tasks, while logicians have pro-
duced an astounding variety of different knowledge representation languages, and
cognitive scientists moved their interests towards dynamical systems and specialized
biologically-inspired simulations. The lone hold-out seemed to be the Cyc project,
which continued to pursue the task of formalizing all 'common-sense' knowledge
in a single knowledge representation language (Lenat 1990). In one critique of
Cyc, Smith instead asked what lessons knowledge representation languages could
learn from hypertext, “Forget intelligence completely, in other words; take the
project as one of constructing the world's largest hypertext system, with Cyc
functioning as a radically improved (and active) counterpart for the Dewey decimal
system. Such a system might facilitate what numerous projects are struggling to
implement: reliable, content-based searching and indexing schemes for massive
textual databases,” a statement that strangely prefigures not only search engines,
but the revitalization of knowledge representation languages due to the Semantic
Web (1991).

3.2 The Resource Description Framework (RDF)

What makes knowledge representation language on the Web different from classical
knowledge representation? Berners-Lee's early thoughts, as given in the first World
Wide Web Conference in Geneva in 1994, were that “adding semantics to the
Web involves two things: allowing documents which have information in machine-
readable forms, and allowing links to be created with relationship values” (Berners-
Lee 1994b). Having information in “machine-readable forms” requires a knowledge
representation language that has some sort of relatively content-neutral language
for encoding (Berners-Lee 1994b). The parallel to knowledge representation in
artificial intelligence is striking, as it also sought to find one universal encoding,
albeit encoding human-intelligence. The second point, of “allowing links,” means
that the basic model of the Semantic Web will be a reflection of the Web itself:
the Semantic Web consists of connecting resources by links. The Semantic Web is
then easily construed as a descendant of semantic networks from classical artificial
intelligence, where nodes are resources and arcs are links. Under the aegis of
the W3C, the first knowledge representation language for the Semantic Web, the
Resource Description Language (RDF) was made a W3C Recommendation, and
it is clearly inspired by work in AI on semantic networks. This should come as
no surprise, for RDF was heavily inspired by the work of R.V. Guha on the Meta-
Content Framework (MCF) (Guha 1996). Before working on MCF, Guha was chief