edu.stanford.nlp.dcoref
Class Document

java.lang.Object
  edu.stanford.nlp.dcoref.Document

All Implemented Interfaces:

java.io.Serializable

public class Document
extends java.lang.Object
implements java.io.Serializable

See Also:

Serialized Form

Nested Class Summary

static class

Document.DocType

Field Summary

java.util.Map<java.lang.Integer,Mention>	allGoldMentions
java.util.Map<java.lang.Integer,Mention>	allPredictedMentions All mentions in a document mentionID -> mention
Annotation	annotation Document annotation
CoNLL2011DocumentReader.Document	conllDoc for conll shared task 2011
java.util.Map<java.lang.Integer,CorefCluster>	corefClusters Clusters for coreferent mentions
Document.DocType	docType The type of document: conversational or article
java.util.Map<java.lang.Integer,CorefCluster>	goldCorefClusters Gold Clusters for coreferent mentions
java.util.List<java.util.List<Mention>>	goldOrderedMentionsBySentence The list of gold mentions
java.util.Set<Pair<java.lang.Integer,java.lang.Integer>>	incompatibles Set of incompatible mention pairs
int	maxUtter
int	numParagraph
int	numSentences
java.util.HashMap<Mention,IntTuple>	positions Position of each mention in the input matrix Each mention occurrence with sentence # and position within sentence (Nth mention, not Nth token)
java.util.List<java.util.List<Mention>>	predictedOrderedMentionsBySentence The list of predicted mentions
java.util.Set<Mention>	roleSet Set of roles (in role apposition) in a document
java.util.Set<Pair<java.lang.Integer,java.lang.Integer>>	speakerPairs mention ID pair
java.util.Map<java.lang.Integer,java.lang.String>	speakers UtteranceAnnotation -> String (speaker): mention ID or speaker string

Constructor Summary

Document()

Document(Annotation anno, java.util.List<java.util.List<Mention>> predictedMentions, java.util.List<java.util.List<Mention>> goldMentions, Dictionaries dict)

Method Summary

protected void	assignOriginalID() When there is no mentionID information (without gold annotation), assign mention IDs
void	extractGoldCorefClusters() Extract gold coref cluster information
protected void	extractGoldLinks() Extract gold coref link information
protected java.util.List<Pair<IntTuple,IntTuple>>	getGoldLinks()
java.util.List<java.util.List<Mention>>	getOrderedMentions() return the list of predicted mentions
static boolean	isSpeaker(Mention m, Mention ant, Dictionaries dict) Check one mention is the speaker of the other mention

Methods inherited from class java.lang.Object

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

Field Detail

docType

public Document.DocType docType

The type of document: conversational or article

annotation

public Annotation annotation

Document annotation

conllDoc

public CoNLL2011DocumentReader.Document conllDoc

for conll shared task 2011

goldOrderedMentionsBySentence

public java.util.List<java.util.List<Mention>> goldOrderedMentionsBySentence

The list of gold mentions

predictedOrderedMentionsBySentence

public java.util.List<java.util.List<Mention>> predictedOrderedMentionsBySentence

The list of predicted mentions

corefClusters

public java.util.Map<java.lang.Integer,CorefCluster> corefClusters

Clusters for coreferent mentions

goldCorefClusters

public java.util.Map<java.lang.Integer,CorefCluster> goldCorefClusters

Gold Clusters for coreferent mentions

allPredictedMentions

public java.util.Map<java.lang.Integer,Mention> allPredictedMentions

All mentions in a document mentionID -> mention

allGoldMentions

public java.util.Map<java.lang.Integer,Mention> allGoldMentions

roleSet

public java.util.Set<Mention> roleSet

Set of roles (in role apposition) in a document

positions

public java.util.HashMap<Mention,IntTuple> positions

Position of each mention in the input matrix Each mention occurrence with sentence # and position within sentence (Nth mention, not Nth token)

speakers

public java.util.Map<java.lang.Integer,java.lang.String> speakers

UtteranceAnnotation -> String (speaker): mention ID or speaker string

speakerPairs

public java.util.Set<Pair<java.lang.Integer,java.lang.Integer>> speakerPairs

mention ID pair

maxUtter

public int maxUtter

numParagraph

public int numParagraph

numSentences

public int numSentences

incompatibles

public java.util.Set<Pair<java.lang.Integer,java.lang.Integer>> incompatibles

Set of incompatible mention pairs

Constructor Detail

Document

public Document()

Document

public Document(Annotation anno,
                java.util.List<java.util.List<Mention>> predictedMentions,
                java.util.List<java.util.List<Mention>> goldMentions,
                Dictionaries dict)

Method Detail

getOrderedMentions

public java.util.List<java.util.List<Mention>> getOrderedMentions()

return the list of predicted mentions

assignOriginalID

protected void assignOriginalID()

When there is no mentionID information (without gold annotation), assign mention IDs

extractGoldCorefClusters

public void extractGoldCorefClusters()

Extract gold coref cluster information

getGoldLinks

protected java.util.List<Pair<IntTuple,IntTuple>> getGoldLinks()

extractGoldLinks

protected void extractGoldLinks()

Extract gold coref link information

isSpeaker

public static boolean isSpeaker(Mention m,
                                Mention ant,
                                Dictionaries dict)

Check one mention is the speaker of the other mention