![]()
![]()
|
Glossary Term:
Definition:
Resource Description Framework (RDF) provides the semantics (meaning) and XML is the syntax used to exchange RDF Meta Data between application programs. A Resource in the Semantic Web is something located at an address, a UniformResourceIdentifier? (URI), or UniformResourceLocator? (URL?). One Resource is related to another by a three-part statement (subject-predicate-object) called an RDF triple. All three parts of the statement can themselves be Resources, located at URIs. The object may also be a text literal. By following the chain of statements, information about a Resource (like a web page) can be extracted. This allows computer programs to read a web page and make sense of its contents beyond just analyzing its text. Search engines of the future will be infinitely more powerful if they can understand (draw inferences about) web page content. It is these inferences that are analogized to how humans find meaning by drawing conclusions. Thus Berners-Lee's dubbing it the Semantic Web. Semantic HTML refers to the HTML tags that describe structure (with implicit meaning) such as h1, blockquote, p, em, and strong, as opposed to those that are purely markup for style like italic, bold, and font. Note that RDF allows relations to be created between Resources (at URIs) that were not imagined when the Resource was created. And note that the relations can be anywhere on the web. In this sense the Semantic Web is to the ordinary web as a RelationalDatabase? is to a HierarchicalDatabase?, much more powerful and flexible. Effectively the Semantic Web is the Web as we know it put into database form and with added metadata. Two foundational standards, the Resource Description Framework (RDF) and the OWL Web Ontology Language (OWL), became W3C recommendations in February.
"There was a lot of pain and sweat and tears and discussions and arguments" in getting RDF and OWL to the level of accepted standards, Berners-Lee? said at W3C 2004. He expects phase two, now in progress, to be more fun: "I hope it'll be very exciting. We'll start to get more satisfaction back from actually building applications and seeing them connect together." Mac World A Semantic Network adds meaning to a SemanticObject? (a node in the network) by linking it to another SemanticObject? (another node). The links are like RDF statements. The nodes are like Resources. A Topic Map consists of Topics (again, Resources), which are linked together by Associations (again links like RDF Statements, but perhaps not directional??). Topics also are linked to Occurrences (of Topics). These might be web pages that mention the Topic. If the Occurrences are regarded as Resources with a relationship of "aboutness" to the Topic Resource, then Topic Map are just special cases of RDF information structures. The Semantic Web is an Ontology (or a web of interdependent ontologies) that uses a Web Ontology Language (OWL) to define all objects (resources) at a URL? or URI by means of statements about their relations to other resources. The goal of the Semantic Web is to allow construction of an InferenceEngine? that can follow the relationships and draw inferences (gain knowledge) about any given resource. It is in this respect the child of fifty years of Artificial Intelligence research. References: Tim Berners-Lee 2001 article in Scientific American W3C Consortium Page on Semantic Web W3C RDF & SW Primer W3C Semantic Web Roadmap Aaron Swartz on SW Purl.org on SW Another Semantic Web Primer XML.com article Infomesh article Dave Beckett on RDF IA Glossary Index | Back |
||||