In Jena the primary use of inference is to generate additional entailments from a set of RDF data. These entailments just appear as additional RDF data in the inferred model and are accessed through the normal API. For example, if an inference engine can determine the class of a resource "foo" is "fooClass" then all Model API calls such as listStatements and getProperty should act as if the triple:
foo rdf:type fooClass .were in the data.
A few reasoner services cannot be made directly available in this way and the InfGraph extension gives access to these - specifically access to validation/consistency checking, derivation traces and find-with-posits.
Note that this interface, and especially the interface onto ValidityReports and Derivations are not yet stable.
| |