Package edu.tufts.hrilab.slug.refResolution

Class Growler

java.lang.Object

edu.tufts.hrilab.slug.refResolution.Growler

public class Growler
extends Object

Growler: Givenness and Relevance Theoretic Open-World Entity Resolution

Field Summary

Fields

Modifier and Type	Field	Description
static final double	DEFAULT_RELEVANCE
static final double	PROBABILITY_THRESHOLD
static final double	RELEVANCE_THRESHOLD
static final double[]	RELEVANCE_WEIGHTINGS

Constructor Summary

Constructors

Constructor	Description
Growler(Resolver resolver, edu.tufts.hrilab.slug.refResolution.ReferenceResolutionComponent.GHSelfModel gh, Map<Symbol,EntityScore> relevanceMap)

Method Summary

Modifier and Type	Method	Description
static Hypothesis	cmclToHypothesis(CrossMappingCandidateList c)
List<Symbol>	domain(String mnemonicAction)
static List<CrossMappingCandidateList>	generateCrossMappingCandidateList(SingleVarCandidateList svc)
List<CrossMappingCandidateList>	generateFullTable(List<SingleVarCandidateList> svcls, List<Term> allTerms)
void	generateInitialValidCandidateList(SingleVarCandidateList svcl, List<Term> allTerms)
LinkedList<String>	getMnemonicActions(Symbol ghTier)
double	getProbabilityOfSatisfyingPolyadicProperties(List<RelevanceTheoreticBinding> bindings, List<Term> allTerms)
double	getProbabilityOfSatisfyingProperties(List<RelevanceTheoreticBinding> relevantBindings, List<Property> props)
double	getProbabilityofSatisfyingUnaryPredicates(RelevanceTheoreticBinding binding, List<Term> allTerms)
double	getRelevance(Symbol ref)
static List<CrossMappingCandidateList>	mergeCrossMappingCandidateLists(List<CrossMappingCandidateList> first, List<CrossMappingCandidateList> second)
List<Hypothesis>	posit_to_LTM(List<Term> terms, List<Variable> varNames, Hypothesis initialHyp)
List<Hypothesis>	resolve_clause(List<Term> semantics, Map<Variable,Symbol> statuses)	The GROWLER algorithm.
List<Hypothesis>	resolve_from_LTM(List<Term> terms, List<Variable> vars, Hypothesis initialHyp, boolean posit)
static double	scaleRelevance(double x)
static double	weightRelevance(EntityScore score)

Methods inherited from class java.lang.Object

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

Field Details

RELEVANCE_THRESHOLD

public static final double RELEVANCE_THRESHOLD

See Also:

• Constant Field Values

PROBABILITY_THRESHOLD

public static final double PROBABILITY_THRESHOLD

See Also:

• Constant Field Values

DEFAULT_RELEVANCE

public static final double DEFAULT_RELEVANCE

See Also:

• Constant Field Values

RELEVANCE_WEIGHTINGS

public static final double[] RELEVANCE_WEIGHTINGS

Constructor Details

Growler

public Growler(Resolver resolver,
 edu.tufts.hrilab.slug.refResolution.ReferenceResolutionComponent.GHSelfModel gh,
 Map<Symbol,EntityScore> relevanceMap)

Method Details

scaleRelevance

public static double scaleRelevance(double x)

weightRelevance

public static double weightRelevance(EntityScore score)

cmclToHypothesis

public static Hypothesis cmclToHypothesis(CrossMappingCandidateList c)

generateCrossMappingCandidateList

public static List<CrossMappingCandidateList> generateCrossMappingCandidateList(SingleVarCandidateList svc)

mergeCrossMappingCandidateLists

public static List<CrossMappingCandidateList> mergeCrossMappingCandidateLists(List<CrossMappingCandidateList> first,
 List<CrossMappingCandidateList> second)

getMnemonicActions

public LinkedList<String> getMnemonicActions(Symbol ghTier)

domain

public List<Symbol> domain(String mnemonicAction)

getRelevance

public double getRelevance(Symbol ref)

getProbabilityofSatisfyingUnaryPredicates

public double getProbabilityofSatisfyingUnaryPredicates(RelevanceTheoreticBinding binding,
 List<Term> allTerms)

getProbabilityOfSatisfyingProperties

public double getProbabilityOfSatisfyingProperties(List<RelevanceTheoreticBinding> relevantBindings,
 List<Property> props)

getProbabilityOfSatisfyingPolyadicProperties

public double getProbabilityOfSatisfyingPolyadicProperties(List<RelevanceTheoreticBinding> bindings,
 List<Term> allTerms)

generateInitialValidCandidateList

public void generateInitialValidCandidateList(SingleVarCandidateList svcl,
 List<Term> allTerms)

generateFullTable

public List<CrossMappingCandidateList> generateFullTable(List<SingleVarCandidateList> svcls,
 List<Term> allTerms)

resolve_clause

public List<Hypothesis> resolve_clause(List<Term> semantics,
 Map<Variable,Symbol> statuses)

The GROWLER algorithm.

Parameters:

semantics - The set of semantic constraints to use during resolution

statuses - The set of status cue mappings for each variable used in those constraints.

Returns:

resolve_from_LTM

public List<Hypothesis> resolve_from_LTM(List<Term> terms,
 List<Variable> vars,
 Hypothesis initialHyp,
 boolean posit)

posit_to_LTM

public List<Hypothesis> posit_to_LTM(List<Term> terms,
 List<Variable> varNames,
 Hypothesis initialHyp)