Class TuplesRankingReasoner

  • All Implemented Interfaces:
    net.sf.tweety.commons.ModelProvider<net.sf.tweety.arg.dung.syntax.Argument,​net.sf.tweety.arg.dung.syntax.DungTheory,​LatticeArgumentRanking>, net.sf.tweety.commons.postulates.PostulateEvaluatable<net.sf.tweety.arg.dung.syntax.Argument>

    public class TuplesRankingReasoner
    extends AbstractRankingReasoner<LatticeArgumentRanking>
    This class implements the "tuples*" argument ranking approach as proposed by [Cayrol, Lagasquie-Schiex. Graduality in argumentation. 2005]. It takes into account all the ancestors branches of an arguument stored in tupled values. Some arguments are incomparable using this approach, i.e. it generates partial rankings.

    Note: This implementation only works for acyclic argument graphs. For cyclic graphs null is returned.
    Author:
    Anna Gessler
    • Constructor Detail

      • TuplesRankingReasoner

        public TuplesRankingReasoner()
    • Method Detail

      • computeTupledValue

        public net.sf.tweety.commons.util.Pair<int[],​int[]> computeTupledValue​(net.sf.tweety.arg.dung.syntax.Argument a,
                                                                                     net.sf.tweety.arg.dung.syntax.DungTheory kb)
        Computes the tupled value for the given argument.
        Parameters:
        a - an argument
        kb - DungTheory
        Returns:
        a pair that consists of the two tuples that represent the defense (first tuple) and attack (second tuple) branches of a.
      • getTupledValues

        public Map<net.sf.tweety.arg.dung.syntax.Argument,​net.sf.tweety.commons.util.Pair<int[],​int[]>> getTupledValues()
        Returns:
        the tupled values computed by previous calls of getModel oder getModels
      • prettyPrintTupledValues

        public String prettyPrintTupledValues()
        Prints the tupled values computed by previous calls of getModel oder getModels in a human-readable way.
        Returns:
        a string representation of the tuples