org.apache.ctakes.coreference.ae

Class MentionClusterRankingCoreferenceAnnotator

java.lang.Object

org.apache.uima.analysis_component.AnalysisComponent_ImplBase

org.apache.uima.analysis_component.Annotator_ImplBase

org.apache.uima.analysis_component.JCasAnnotator_ImplBase

org.apache.uima.fit.component.JCasAnnotator_ImplBase

org.cleartk.ml.CleartkAnnotator<Double>

org.apache.ctakes.coreference.ae.MentionClusterRankingCoreferenceAnnotator

All Implemented Interfaces:

org.apache.uima.analysis_component.AnalysisComponent, org.apache.uima.fit.factory.initializable.Initializable

@PipeBitInfo(name="Coreference (Cluster Rank)",
             description="Coreference annotator using mention-synchronous paradigm.",
             dependencies={BASE_TOKEN,SENTENCE,SECTION,PARAGRAPH,IDENTIFIED_ANNOTATION,MARKABLE},
             products=COREFERENCE_RELATION)
public class MentionClusterRankingCoreferenceAnnotator
extends org.cleartk.ml.CleartkAnnotator<Double>

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair

Field Summary

Fields

Modifier and Type	Field and Description
static String	CLUSTER_RELATION_CATEGORY
protected Random	coin
static String	NO_RELATION_CATEGORY
static String	PARAM_PROBABILITY_OF_KEEPING_A_NEGATIVE_EXAMPLE
protected double	probabilityOfKeepingANegativeExample

Fields inherited from class org.cleartk.ml.CleartkAnnotator

classifier, dataWriter, PARAM_CLASSIFIER_FACTORY_CLASS_NAME, PARAM_DATA_WRITER_FACTORY_CLASS_NAME, PARAM_IS_TRAINING

Constructor Summary

Constructors

Constructor and Description
MentionClusterRankingCoreferenceAnnotator()

Method Summary

Modifier and Type	Method and Description
protected Double	classify(List<org.cleartk.ml.Feature> features) Predict an outcome given a set of features.
static org.apache.uima.analysis_engine.AnalysisEngineDescription	createAnnotatorDescription(String modelPath)
static org.apache.uima.analysis_engine.AnalysisEngineDescription	createDataWriterDescription(Class<? extends org.cleartk.ml.DataWriter<?>> dataWriterClass, File outputDirectory, float downsamplingRate)
protected void	createRelation(org.apache.uima.jcas.JCas jCas, CollectionTextRelation cluster, IdentifiedAnnotation mention, String predictedCategory) Create a UIMA relation type based on arguments and the relation label.
Set<String>	getBestEnt(org.apache.uima.jcas.JCas jcas, CollectionTextRelation cluster)
Set<String>	getBestEnt(org.apache.uima.jcas.JCas jcas, Markable markable)
protected Iterable<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair>	getCandidateRelationArgumentPairs(org.apache.uima.jcas.JCas jcas, IdentifiedAnnotation mention)
protected List<RelationFeaturesExtractor<CollectionTextRelation,IdentifiedAnnotation>>	getFeatureExtractors()
protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair>	getHeadwordMatchPairs(org.apache.uima.jcas.JCas jcas, IdentifiedAnnotation mention, int sentDist)
Map<RelationExtractorEvaluation.HashableArguments,Double>	getMarkablePairScores(org.apache.uima.jcas.JCas jCas)
protected List<org.cleartk.ml.feature.extractor.FeatureExtractor1<Markable>>	getMentionExtractors()
protected String	getRelationCategory(Map<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair,CollectionTextRelationIdentifiedAnnotationRelation> relationLookup, CollectionTextRelation cluster, IdentifiedAnnotation mention) Looks up the arguments in the specified lookup table and converts the relation into a label for classification
protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair>	getSectionHeaderPairs(org.apache.uima.jcas.JCas jcas, IdentifiedAnnotation mention, int sentDist)
protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair>	getSentenceDistancePairs(org.apache.uima.jcas.JCas jcas, IdentifiedAnnotation mention, int sentDist)
void	process(org.apache.uima.jcas.JCas jCas)

Methods inherited from class org.cleartk.ml.CleartkAnnotator

collectionProcessComplete, initialize, isTraining

Methods inherited from class org.apache.uima.fit.component.JCasAnnotator_ImplBase

getLogger

Methods inherited from class org.apache.uima.analysis_component.JCasAnnotator_ImplBase

getRequiredCasInterface, process

Methods inherited from class org.apache.uima.analysis_component.Annotator_ImplBase

getCasInstancesRequired, hasNext, next

Methods inherited from class org.apache.uima.analysis_component.AnalysisComponent_ImplBase

batchProcessComplete, destroy, getContext, getResultSpecification, reconfigure, setResultSpecification

Methods inherited from class java.lang.Object

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

Field Detail

NO_RELATION_CATEGORY

public static final String NO_RELATION_CATEGORY

See Also:

Constant Field Values

CLUSTER_RELATION_CATEGORY

public static final String CLUSTER_RELATION_CATEGORY

See Also:

Constant Field Values

PARAM_PROBABILITY_OF_KEEPING_A_NEGATIVE_EXAMPLE

public static final String PARAM_PROBABILITY_OF_KEEPING_A_NEGATIVE_EXAMPLE

See Also:

Constant Field Values

probabilityOfKeepingANegativeExample

protected double probabilityOfKeepingANegativeExample

coin

protected Random coin

Constructor Detail

MentionClusterRankingCoreferenceAnnotator

public MentionClusterRankingCoreferenceAnnotator()

Method Detail

createDataWriterDescription

public static org.apache.uima.analysis_engine.AnalysisEngineDescription createDataWriterDescription(Class<? extends org.cleartk.ml.DataWriter<?>> dataWriterClass,
                                                                                                    File outputDirectory,
                                                                                                    float downsamplingRate)
                                                                                             throws org.apache.uima.resource.ResourceInitializationException

Throws:

org.apache.uima.resource.ResourceInitializationException

createAnnotatorDescription

public static org.apache.uima.analysis_engine.AnalysisEngineDescription createAnnotatorDescription(String modelPath)
                                                                                            throws org.apache.uima.resource.ResourceInitializationException

Throws:

org.apache.uima.resource.ResourceInitializationException

getFeatureExtractors

protected List<RelationFeaturesExtractor<CollectionTextRelation,IdentifiedAnnotation>> getFeatureExtractors()

getMentionExtractors

protected List<org.cleartk.ml.feature.extractor.FeatureExtractor1<Markable>> getMentionExtractors()

getCandidateRelationArgumentPairs

protected Iterable<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair> getCandidateRelationArgumentPairs(org.apache.uima.jcas.JCas jcas,
                                                                                                                                               IdentifiedAnnotation mention)

getSentenceDistancePairs

protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair> getSentenceDistancePairs(org.apache.uima.jcas.JCas jcas,
                                                                                                                                  IdentifiedAnnotation mention,
                                                                                                                                  int sentDist)

getSectionHeaderPairs

protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair> getSectionHeaderPairs(org.apache.uima.jcas.JCas jcas,
                                                                                                                               IdentifiedAnnotation mention,
                                                                                                                               int sentDist)

getHeadwordMatchPairs

protected List<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair> getHeadwordMatchPairs(org.apache.uima.jcas.JCas jcas,
                                                                                                                               IdentifiedAnnotation mention,
                                                                                                                               int sentDist)

process

public void process(org.apache.uima.jcas.JCas jCas)
             throws org.apache.uima.analysis_engine.AnalysisEngineProcessException

Specified by:

process in class org.apache.uima.analysis_component.JCasAnnotator_ImplBase

Throws:

org.apache.uima.analysis_engine.AnalysisEngineProcessException

getRelationCategory

protected String getRelationCategory(Map<MentionClusterRankingCoreferenceAnnotator.CollectionTextRelationIdentifiedAnnotationPair,CollectionTextRelationIdentifiedAnnotationRelation> relationLookup,
                                     CollectionTextRelation cluster,
                                     IdentifiedAnnotation mention)

Looks up the arguments in the specified lookup table and converts the relation into a label for classification

Returns:

If this category should not be processed for training return null otherwise it returns the label sent to the datawriter

classify

protected Double classify(List<org.cleartk.ml.Feature> features)
                   throws org.cleartk.ml.CleartkProcessingException

Predict an outcome given a set of features. By default, this simply delegates to the object's classifier. Subclasses may override this method to implement more complex classification procedures.

Parameters:

features - The features to be classified.

Returns:

The predicted outcome (label) for the features.

Throws:

org.cleartk.ml.CleartkProcessingException

createRelation

protected void createRelation(org.apache.uima.jcas.JCas jCas,
                              CollectionTextRelation cluster,
                              IdentifiedAnnotation mention,
                              String predictedCategory)

Create a UIMA relation type based on arguments and the relation label. This allows subclasses to create/define their own types: e.g. coreference can create CoreferenceRelation instead of BinaryTextRelation

Parameters:

jCas - - JCas object, needed to create new UIMA types

arg1 - - First argument to relation

arg2 - - Second argument to relation

predictedCategory - - Name of relation

getBestEnt

public Set<String> getBestEnt(org.apache.uima.jcas.JCas jcas,
                              CollectionTextRelation cluster)

getBestEnt

public Set<String> getBestEnt(org.apache.uima.jcas.JCas jcas,
                              Markable markable)

getMarkablePairScores

public Map<RelationExtractorEvaluation.HashableArguments,Double> getMarkablePairScores(org.apache.uima.jcas.JCas jCas)