How are .RDF files produced?

0
votes

Consuming a SPARQL endpoint for the first time, I see that there are a number of prefixes that when I copy/paste into a browser result in downloading .rdf files. These files reference rdf, rdfs, and owl namespaces, so I assume that these files represent graph resources according to the RDF, RDFS, and OWL specifications.

What are different ways that these files can be generated? i.e. are they written by hand, are they produced from databases, etc.

The files that I'm using are publicly available at:

The reason I'm asking is that reading this question - What's a RDF triple? - I see that it would be possible to parse a .rdf files down to a list of triples.

1
rdf

1 Answers

2
votes

RDF triples describing an ontology (or the T-box, T for "terminology") can be produced by hand using any text editor. However, they are usually produced with ontology editors.

RDF triples describing concrete resources, which are colloquially referred to as "instance data" (it is not quite correct because an instance is an instance of, so the primacy of classes is assumed; that is a valid assumption for OO but not for RDF) or as A-box (A for "assertion") can be produces through:

  1. Transformation from non-RDF format (e.g. tabular data, using e.g. R2RML)
  2. Inferencing, when new triples a produced due to asserted triples (A), ontology (T) and rules (R), incl. application of logical axioms). The R determines that when certain data A is subject to constraints defined by T, then new triples are inferred describing the consequences of the assertions.
  3. User input, which can be transformed to e.g. SPARQL UPDATE patterns.