I've just read the extensive post by Nova Spivack titled: The Semantic Web, Collective Intelligence and Hyperdata, courtesy of a post by Danny Ayres titled: Confused about the Semantic Web , in response to a post by Tim O'Reilly titled: Economist Confused About the Semantic Web? .

My Comments:

Hyperdata is short for HyperLinked Data :-) The same applies to Linked Data. Thus, we have two literal labels for the same core Concept. HTTP is the enabling protocol for "Hyper-linking" Documents and associated Structured Data via the World Wide Web (Web for short). Data Links associated with Structured Data contained in, or hosted by, Documents on the Web.

RDFa, eRDF, GRDDL, SPARQL Query Language, SPARQL Protocol (SOAP or REST service), SPARQL Results Serializations (XML or JSON) collectively provide a myriad of unobtrusive routes to structured data embedded within, or associated with, existing Web Documents.

As Danny already states, ontologies are not prerequisites for producing structured data using the RDF Data Model. They simply aid the ability to express one's self clearly (i.e. no repetition or ambiguity) across a broad audience of machines (directly) and their human masters (indirectly).

Using the crux of this post as the anecdote: The Semantic Data Web would simplify the process of claiming and/or proving that Linked Data and Hyperdata describe the same concept. It achieves this by using Triples (Subject, Predicate, Object) expressed in various forms (N3, Turtle, RDF/XML etc.) to formalize claims in a form palatable to electronic agents (machines) operating on behalf of Humans. In a nutshell, this increases human productive by completely obliterates the erstwhile exponential costs of discovering data, information, and knowledge.

BTW - for full effect, view this post (i.e. cut and paste the Permalink URI of this post, below) into an RDF Browser such as: