amie.mining.assistant.experimental

Class ExistentialRulesHeadVariablesMiningAssistant

java.lang.Object

amie.mining.assistant.MiningAssistant

amie.mining.assistant.DefaultMiningAssistant

amie.mining.assistant.experimental.ExistentialRulesHeadVariablesMiningAssistant

public class ExistentialRulesHeadVariablesMiningAssistant
extends DefaultMiningAssistant

Extension of the default mining assistant that also reports existential rules, i.e., rules where one of the head variables is allowed to be non-closed.

Author:

galarrag

Field Summary

Fields inherited from class amie.mining.assistant.DefaultMiningAssistant

hardQueries

Fields inherited from class amie.mining.assistant.MiningAssistant

allowConstants, avoidUnboundTypeAtoms, bodyExcludedRelations, bodyTargetRelations, confidenceMetric, countAlwaysOnSubject, enabledConfidenceUpperBounds, enabledFunctionalityHeuristic, enablePerfectRules, enableQueryRewriting, enforceConstants, exploitMaxLengthOption, headCardinalities, headExcludedRelations, kb, kbSchema, maxDepth, minPcaConfidence, minStdConfidence, recursivityLimit, subPropertyString, totalObjectCount, totalSubjectCount, typeString, verbose

Constructor Summary

Constructors

Constructor and Description
ExistentialRulesHeadVariablesMiningAssistant(KB dataSource)

Method Summary

Modifier and Type	Method and Description
void	calculateConfidenceMetrics(Rule candidate) It computes the standard and the PCA confidence of a given rule.
java.lang.String	getDescription() Brief description of the MiningAssistant capabilities.
boolean	testConfidenceThresholds(Rule candidate) It checks whether a rule satisfies the confidence thresholds and the sky-line heuristic: the strategy that avoids outputting rules that do not improve the confidence w.r.t their parents.

Methods inherited from class amie.mining.assistant.DefaultMiningAssistant

computeBodySize, computeCardinality, computePcaBodySize, computePCAConfidence, computeStandardConfidence, containsHardCase, getClosingAtoms, getDanglingAtoms, getDanglingAtoms, getInitialAtoms, getInitialAtomsFromSeeds, getInstantiatedAtoms, getTotalCount, main, rewriteProjectionQuery

Methods inherited from class amie.mining.assistant.MiningAssistant

buildInitialQueries, calculateConfidenceApproximationFor3Atoms, calculateConfidenceApproximationForGeneralCase, calculateConfidenceBoundsAndApproximations, canAddInstantiatedAtoms, findCountingVariable, getBodyExcludedRelations, getBodyTargetRelations, getConfidenceMetric, getFactsCount, getHeadCardinality, getHeadExcludedRelations, getInstantiatedAtoms, getKb, getKbSchema, getMaxDepth, getMinConfidence, getPcaConfidenceThreshold, getRecursivityLimit, getRelationCardinality, getRelationCardinality, getSubClasses, getTotalCount, getTotalObjectCount, getTotalSubjectCount, isAvoidUnboundTypeAtoms, isEnabledConfidenceUpperBounds, isEnabledFunctionalityHeuristic, isEnablePerfectRules, isEnableQueryRewriting, isEnforceConstants, isExploitMaxLengthOption, isVerbose, registerHeadRelation, setAllowConstants, setAvoidUnboundTypeAtoms, setBodyExcludedRelations, setConfidenceMetric, setCountAlwaysOnSubject, setEnabledConfidenceUpperBounds, setEnabledFunctionalityHeuristic, setEnablePerfectRules, setEnableQueryRewriting, setEnforceConstants, setExploitMaxLengthOption, setHeadExcludedRelations, setKbSchema, setMaxDepth, setPcaConfidenceThreshold, setRecursivityLimit, setStdConfidenceThreshold, setTargetBodyRelations, setVerbose, testLength

Methods inherited from class java.lang.Object

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

Constructor Detail

ExistentialRulesHeadVariablesMiningAssistant

public ExistentialRulesHeadVariablesMiningAssistant(KB dataSource)

Method Detail

getDescription

public java.lang.String getDescription()

Description copied from class: MiningAssistant

Brief description of the MiningAssistant capabilities.

Overrides:

getDescription in class DefaultMiningAssistant

calculateConfidenceMetrics

public void calculateConfidenceMetrics(Rule candidate)

Description copied from class: MiningAssistant

It computes the standard and the PCA confidence of a given rule. It assumes the rule's cardinality (absolute support) is known.

Overrides:

calculateConfidenceMetrics in class DefaultMiningAssistant

testConfidenceThresholds

public boolean testConfidenceThresholds(Rule candidate)

Description copied from class: MiningAssistant

It checks whether a rule satisfies the confidence thresholds and the sky-line heuristic: the strategy that avoids outputting rules that do not improve the confidence w.r.t their parents.

Overrides:

testConfidenceThresholds in class MiningAssistant

Returns: