Resource Description Framework data consists of resources (nodes), and property/value pairs describing the resource. A node is any object which can be pointed to by a URI, properties are attributes of the node, and values can be either atomic values for the attribute, or other nodes. For example, information about a particular web page (a node), might include the property "Author". The value for the Author property could be either a string giving the name of the author, or a link to a resource describing the author.
Resource Description Framework only specifies a mechanism for encoding and transferring metadata. It does not specify what that metadata should, or can be. RDF does not, for example, define an "Author" attribute. Sets of properties are defined within RDF Vocabularies (or Schemas). Anynone can create an RDF schema, describing a specialized set of properties, by creating a resource, referenced by the Schema URI, which provides a human- and machine-understandable definition of the schema's properties. The description of a node may include properties defined in different schemas. The properties within a resource description are associated with a certain schema definition using the XML namespace mechanism. Schemas currently being developed include a content screening system modeled after PICS, and a bibliographic vocabulary, such as the Dublin Core Initiative.