Package org.tweetyproject.arg.dung.learning

Class SimpleAFLearner

java.lang.Object

org.tweetyproject.arg.dung.learning.SimpleAFLearner

All Implemented Interfaces:

AFLearner

public class SimpleAFLearner
extends Object
implements AFLearner

Naive implementation of the iterative algorithm for learning argumentation frameworks from labelings

Author:

Lars Bengel

Constructor Summary

Constructors

Constructor	Description
SimpleAFLearner(Collection<Argument> args)	initialize the Learner with a set of arguments

Method Summary

Modifier and Type	Method	Description
Map<Argument,Collection<Collection<Attack>>>	computePartialAttackRelations()	TODO is this still necessary?
DungTheory	getModel()	compute an argumentation framework that satisfy the internal acceptance conditions
Collection<DungTheory>	getModels()	compute the set of argumentation frameworks that satisfy all acceptance conditions 1.
long	getNumberOfFrameworks()	compute the number of argumentation frameworks that satisfy the internal acceptance conditions
long	getNumberOfFrameworks(boolean shortcut)	computes whether there is more than one argumentation framework that satisfies the internal conditions helpful for break conditions when learning
boolean	learnLabeling(Input labeling)	learn a single labeling i.e.
void	printStatus()	print the internal acceptance conditions
void	printStatus(boolean dnf)

Methods inherited from class java.lang.Object

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

Constructor Details

SimpleAFLearner

public SimpleAFLearner(Collection<Argument> args)

initialize the Learner with a set of arguments

Parameters:

args - a set of arguments

Method Details

learnLabeling

public boolean learnLabeling(Input labeling)

learn a single labeling i.e. compute the acceptance condition for each argument a wrt to the given labeling and combine it with the previous acceptance condition of a

Specified by:

learnLabeling in interface AFLearner

Parameters:

labeling - some labeling of the set of arguments

Returns:

true if the labeling was learned successfully

getNumberOfFrameworks

public long getNumberOfFrameworks()

Description copied from interface: AFLearner

compute the number of argumentation frameworks that satisfy the internal acceptance conditions

Specified by:

getNumberOfFrameworks in interface AFLearner

Returns:

number of afs that produce all processed input labelings

getNumberOfFrameworks

public long getNumberOfFrameworks(boolean shortcut)

computes whether there is more than one argumentation framework that satisfies the internal conditions helpful for break conditions when learning

Specified by:

getNumberOfFrameworks in interface AFLearner

Parameters:

shortcut - if true, then only return 1 or -1

Returns:

1 if there is only one AF, -1 if there is more than one

computePartialAttackRelations

public Map<Argument,Collection<Collection<Attack>>> computePartialAttackRelations()

TODO is this still necessary?

Returns:

the partial attack relations

getModels

public Collection<DungTheory> getModels()

compute the set of argumentation frameworks that satisfy all acceptance conditions 1. for each acceptance condition (argument) compute all models 2. for each model compute a corresponding attack relation 3. compute all possible combinations of the attack relations and we have all argumentation frame works satisfying the conditions

Specified by:

getModels in interface AFLearner

Returns:

the set of argumentation frameworks obtained from learning

getModel

public DungTheory getModel()

Description copied from interface: AFLearner

compute an argumentation framework that satisfy the internal acceptance conditions

Specified by:

getModel in interface AFLearner

Returns:

some argumentation framework that satisfies the internal conditions

printStatus

public void printStatus()

Description copied from interface: AFLearner

print the internal acceptance conditions

Specified by:

printStatus in interface AFLearner

printStatus

public void printStatus(boolean dnf)

Parameters:

dnf - if true, the acceptance conditions are transformed to DNF before printing print the current status of the acceptance conditions