The W3C introduction to the Resource Description Framework — triples, IRIs, literals, blank nodes, triple terms, named graphs, seven serialization formats, and RDF semantics. New in 1.2: directional language-tagged strings and triple terms for statement reification.
RDF expresses information as triples: <subject> <predicate> <object>. A set of triples forms a graph. Multiple named graphs plus a default graph form an RDF dataset.
| Format | Graphs | Key Features |
|---|---|---|
| Turtle | Single | Prefixes, semicolons, a shorthand, [] for blank nodes, <<>> and {||} for reification |
| TriG | Multiple | GRAPH blocks extend Turtle for named graphs |
| N-Triples | Single | Line-based, full IRIs — for bulk exchange |
| N-Quads | Multiple | Fourth element per line = graph IRI |
| JSON-LD | Multiple | JSON syntax with @context, @id, @type, @reverse |
| RDFa | Single | Embedded in HTML/XML — resource, property, typeof |
| RDF/XML | Single | Original XML syntax — rdf:Description, rdf:about |
Triple terms enable statements about statements — annotate assertions with provenance or context without implying truth. New in RDF 1.2.
New 1.2RDFS reasoners deduce additional triples from ground statements. Domain/range axioms derive type relationships automatically.
FOAF, Dublin Core, schema.org, SKOS provide shared vocabularies. Vocabularies get their value from reuse. owl:sameAs links equivalents.
Group triples by source in named graphs. TriG and N-Quads provide syntax. Graph names often indicate provenance or data context.
Publishing interlinked structured data on the Web. Wikidata, DBpedia, Europeana, VIAF are examples. Retrieving one IRI yields data linking to others.
Language-tagged strings now support ltr and rtl base direction. Proper bidirectional text rendering for Arabic, Hebrew, and mixed-script content. New in RDF 1.2.