triples

Type Members

1. abstract class AbstractForwardRuleReasoner[Rdf <: RDF, D, G <: AbstractRDFGraphSpark[Rdf, D, G]] extends Logging

   

   A forward chaining based reasoner.

2. abstract class AbstractForwardRuleReasonerRDFS[Rdf <: RDF, D, G <: AbstractRDFGraphSpark[Rdf, D, G]] extends TransitiveReasoner

   

   A forward chaining implementation of the RDFS entailment regime.

3. trait FixpointIteration[T] extends Logging

   

   Creates a new RDD by performing bulk iterations using the given step function.

   Creates a new RDD by performing bulk iterations using the given step function. The first RDD the step function returns is the input for the next iteration, the second RDD is the termination criterion. The iterations terminate when either the termination criterion RDD contains no elements or when maxIterations iterations have been performed.

4. trait ForwardRuleReasoner extends Profiler

   

   A forward chaining based reasoner.

5. class ForwardRuleReasonerEL extends TransitiveReasoner

   

   This rule-based forward chaining reasoner implementation is based on the paper

   This rule-based forward chaining reasoner implementation is based on the paper

   'Pushing the EL envelope' by Baader, Brandt, Lutz. IJCAI. Vol. 5. 2005.

   The paper does not cover the whole OWL 2 EL profile and makes some assumptions w.r.t. the allowed axioms. Given the set BC that contains

   - owl:Thing - all concept names - all nominals { a } (with just one individual!) - concrete domain descriptions p(f1, ..., fk) (not used here)

   The allowed general concept inclusions are the following (with C1, C2 \in BC, D \in BC \cup { owl:Nothin }):

   - C1 \sqsubseteq D - C1 \sqcap C2 \sqsubseteq D - C1 \sqsubseteq \exists r.C2 - \exists r.C1 \sqsubseteq D

   Role inclusions may be of the form

   - r \sqsubseteq s , or - r1 o r2 \sqsubseteq s

   We implemented the completion rules that were feasible in our setting.

   Naming conventions: - C, C1, C2 --> class names, or nominals (with just one individual), or owl:Thing - D, D1, D2 --> class names, or nominals (with just one individual), owl:Thing, or owl:Nothing

6. class ForwardRuleReasonerNaive extends AbstractForwardRuleReasoner[Jena, RDD[Triple], RDFGraphNative]

   

   A naive implementation of the forward chaining based reasoner that does fix-point iteration, i.e.

   A naive implementation of the forward chaining based reasoner that does fix-point iteration, i.e. it applies all rules in each iteration until no new data has been generated. .

7. class ForwardRuleReasonerOWLHorst extends TransitiveReasoner

   

   A forward chaining implementation of the OWL Horst entailment regime.

8. abstract class ForwardRuleReasonerOptimized[Rdf <: RDF, D, N <: triples.ForwardRuleReasonerOptimized.Rdf.Node, T <: triples.ForwardRuleReasonerOptimized.Rdf.Triple, G <: AbstractRDFGraphSpark[Rdf, D, G]] extends AbstractForwardRuleReasoner[Rdf, D, G]

   

   An optimized implementation of the forward chaining based reasoner.

9. class ForwardRuleReasonerOptimizedNative extends ForwardRuleReasonerOptimized[Jena, RDD[Triple], Node, Triple, RDFGraphNative]

   

   An optimized implementation of the forward chaining based reasoner using Spark data structures and operations.

10. class ForwardRuleReasonerRDFS extends TransitiveReasoner

    

    A forward chaining implementation of the RDFS entailment regime.

11. class ForwardRuleReasonerRDFSDataframe extends TransitiveReasoner

    

    A forward chaining implementation of the RDFS entailment regime.

12. class TransitiveReasoner extends ForwardRuleReasoner

    

    An engine to compute the transitive closure (TC) for a set of triples given in several datastructures.