Package net.sf.tweety.arg.rankings.reasoner

Class CategorizerRankingReasoner

java.lang.Object

net.sf.tweety.arg.rankings.reasoner.AbstractRankingReasoner<NumericalArgumentRanking>

net.sf.tweety.arg.rankings.reasoner.CategorizerRankingReasoner

All Implemented Interfaces:

ModelProvider<Argument,DungTheory,NumericalArgumentRanking>, PostulateEvaluatable<Argument>

public class CategorizerRankingReasoner
extends AbstractRankingReasoner<NumericalArgumentRanking>

This class implements the "h-categorizer" argument ranking approach that was originally proposed by [Besnard, Hunter. A logic-based theory of deductive arguments. 2001] for deductive logics. It uses the Fixed-point algorithm of [Pu, Zhang, Luo, Luo. Argument Ranking with Categoriser Function. KSEM 2014] which allows for cycles in argumentation graphs.

Author:

Anna Gessler

See Also:

HCategorizer

Constructor Summary

Constructors

Constructor	Description
CategorizerRankingReasoner()

Method Summary

Modifier and Type	Method	Description
private double	calculateCategorizerFunction(double[] v_old, Matrix directAttackMatrix, int i)	Computes the h-Categorizer function.
private double	getDistance(double[] v_old, double[] v)	Computes the Euclidean distance between to the given arrays.
NumericalArgumentRanking	getModel(DungTheory base)	Returns a single (dedicated) model of the given belief base.
java.util.Collection<NumericalArgumentRanking>	getModels(DungTheory bbase)	Returns a characterizing model of the given belief base

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

CategorizerRankingReasoner

public CategorizerRankingReasoner()

Method Detail

getModels

public java.util.Collection<NumericalArgumentRanking> getModels(DungTheory bbase)

Description copied from interface: ModelProvider

Returns a characterizing model of the given belief base

Parameters:

bbase - some belief base

Returns:

the (selected) models of the belief base

getModel

public NumericalArgumentRanking getModel(DungTheory base)

Description copied from interface: ModelProvider

Returns a single (dedicated) model of the given belief base. If the implemented method allows for more than one dedicated model, the selection may be non-deterministic.

Parameters:

base - some belief base

Returns:

a selected model of the belief base.

calculateCategorizerFunction

private double calculateCategorizerFunction(double[] v_old,
                                            Matrix directAttackMatrix,
                                            int i)

Computes the h-Categorizer function.

Parameters:

v_old - array of double valuations that were computed in the previous iteration

directAttackMatrix - complete matrix of direct attacks

i - row of the attack matrix that will be used in the calculation

Returns:

categorizer valuation

getDistance

private double getDistance(double[] v_old,
                           double[] v)

Computes the Euclidean distance between to the given arrays.

Parameters:

v_old - first array

v - second array

Returns:

distance between v and v_old