This is an automated email from the ASF dual-hosted git repository.
mawiesne pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/opennlp-sandbox.git
The following commit(s) were added to refs/heads/main by this push:
new 7544e3c Extract constants, symbols and enhance CoRef JavaDoc (#318)
7544e3c is described below
commit 7544e3ce67102acbd6402941894aff5a1fbe19ec
Author: Martin Wiesner <[email protected]>
AuthorDate: Wed Jul 23 10:32:46 2025 +0200
Extract constants, symbols and enhance CoRef JavaDoc (#318)
extracts constants of frequently used POS tag strings
extracts constants of model names
simplifies code in AbstractLinker
adds missing @Override annotations
enhances JavaDoc along the path
fixes typos
---
.../cmdline/coref/CoreferenceConverterTool.java | 1 -
.../tools/cmdline/coref/TrainingParams.java | 2 +
.../java/opennlp/tools/coref/DiscourseElement.java | 2 +-
.../opennlp/tools/coref/linker/AbstractLinker.java | 48 +++++++++++-----------
.../opennlp/tools/coref/linker/DefaultLinker.java | 18 ++++----
.../java/opennlp/tools/coref/linker/Linker.java | 7 +---
.../opennlp/tools/coref/linker/LinkerMode.java | 6 ++-
.../opennlp/tools/coref/mention/HeadFinder.java | 14 ++++---
.../java/opennlp/tools/coref/mention/Mention.java | 5 +--
.../tools/coref/mention/MentionContext.java | 4 +-
.../opennlp/tools/coref/mention/MentionFinder.java | 29 +++++++------
.../tools/coref/mention/PTBMentionFinder.java | 14 +++----
.../java/opennlp/tools/coref/mention/Parse.java | 8 ++--
.../coref/mention/ShallowParseMentionFinder.java | 7 +++-
.../tools/coref/resolver/AbstractResolver.java | 3 +-
.../tools/coref/resolver/CommonNounResolver.java | 12 +++---
.../resolver/DefaultNonReferentialResolver.java | 12 +++---
.../tools/coref/resolver/DefiniteNounResolver.java | 6 ++-
.../opennlp/tools/coref/resolver/IsAResolver.java | 2 +-
.../tools/coref/resolver/MaxentResolver.java | 10 +++--
.../coref/resolver/NonReferentialResolver.java | 10 +++--
.../tools/coref/resolver/PerfectResolver.java | 10 ++---
.../tools/coref/resolver/PluralNounResolver.java | 12 +++---
.../coref/resolver/PluralPronounResolver.java | 8 ++--
.../tools/coref/resolver/ProperNounResolver.java | 10 +++--
.../opennlp/tools/coref/resolver/Resolver.java | 5 +++
.../tools/coref/resolver/ResolverUtils.java | 8 ++--
.../coref/resolver/SingularPronounResolver.java | 14 ++++---
.../coref/resolver/SpeechPronounResolver.java | 24 ++++++-----
.../main/java/opennlp/tools/coref/sim/Context.java | 9 ++--
.../main/java/opennlp/tools/coref/sim/Gender.java | 2 +
.../main/java/opennlp/tools/coref/sim/Number.java | 2 +
.../java/opennlp/tools/coref/sim/TrainModel.java | 2 +-
.../tools/formats/muc/MucCorefContentHandler.java | 18 ++++----
.../tools/lang/english/TreebankNameFinder.java | 22 ++++++----
.../tools/coref/linker/AbstractLinkerTest.java | 2 +-
36 files changed, 208 insertions(+), 160 deletions(-)
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/CoreferenceConverterTool.java
b/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/CoreferenceConverterTool.java
index 4920d18..512a7ea 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/CoreferenceConverterTool.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/CoreferenceConverterTool.java
@@ -20,7 +20,6 @@ package opennlp.tools.cmdline.coref;
import opennlp.tools.cmdline.AbstractConverterTool;
import opennlp.tools.coref.CorefSample;
import opennlp.tools.coref.CorefSampleStreamFactory;
-import opennlp.tools.postag.POSSample;
/**
* Tool to convert multiple data formats into native OpenNLP Coref training
format.
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/TrainingParams.java
b/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/TrainingParams.java
index efd9034..c54104d 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/TrainingParams.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/cmdline/coref/TrainingParams.java
@@ -18,11 +18,13 @@
package opennlp.tools.cmdline.coref;
import opennlp.tools.cmdline.params.BasicTrainingParams;
+import opennlp.tools.commons.Internal;
/**
* TrainingParameters for Co-Referencer
* <p>
* Note: Do not use this class, internal use only!
*/
+@Internal
interface TrainingParams extends BasicTrainingParams {
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/DiscourseElement.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/DiscourseElement.java
index a74ae01..c7ecc81 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/DiscourseElement.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/DiscourseElement.java
@@ -48,7 +48,7 @@ public abstract class DiscourseElement {
/**
* @return Retrieves an {@link Iterator} over the mentions which iterates
through them
- * based on which were most recently mentioned..
+ * based on which were most recently mentioned.
*/
public Iterator<MentionContext> getRecentMentions() {
return new ReverseListIterator<>(extents);
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/AbstractLinker.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/AbstractLinker.java
index aae7cb3..5b020c1 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/AbstractLinker.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/AbstractLinker.java
@@ -19,6 +19,9 @@ package opennlp.tools.coref.linker;
import java.io.IOException;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
import opennlp.tools.coref.DiscourseEntity;
import opennlp.tools.coref.DiscourseModel;
import opennlp.tools.coref.mention.HeadFinder;
@@ -28,10 +31,9 @@ import opennlp.tools.coref.mention.MentionFinder;
import opennlp.tools.coref.mention.Parse;
import opennlp.tools.coref.resolver.AbstractResolver;
import opennlp.tools.coref.sim.Gender;
+import opennlp.tools.coref.sim.GenderEnum;
import opennlp.tools.coref.sim.Number;
-
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
+import opennlp.tools.coref.sim.NumberEnum;
/**
* Provides a default implementation of many of the methods in {@link Linker}
that
@@ -151,43 +153,43 @@ public abstract class AbstractLinker implements Linker {
}
/**
- * Updates the specified {@link DiscourseModel} with the specified mention
as coreferent with the specified e.
+ * Updates the specified {@link DiscourseModel} with the specified mention
as coreferent
+ * with the specified {@code entity 'e'}.
+ *
* @param dm The {@link DiscourseModel}.
* @param m The {@link MentionContext mention} to be added to the specified
{@code entity 'e'}.
* @param e The {@link DiscourseEntity} which is mentioned by the specified
mention.
* @param useDiscourseModel Whether the mentions should be kept as an e or
simply co-indexed.
*/
protected void updateExtent(DiscourseModel dm, MentionContext m,
DiscourseEntity e, boolean useDiscourseModel) {
+ final NumberEnum n = m.getNumber();
+ final GenderEnum g = m.getGender();
+ final double gProb = m.getGenderProb();
+ final double nProb = m.getNumberProb();
if (useDiscourseModel) {
if (e != null) {
logger.debug("Adding extent: {}", m.toText());
- if (e.getGenderProbability() < m.getGenderProb()) {
- e.setGender(m.getGender());
- e.setGenderProbability(m.getGenderProb());
+ if (e.getGenderProbability() < gProb) {
+ e.setGender(g);
+ e.setGenderProbability(gProb);
}
- if (e.getNumberProbability() < m.getNumberProb()) {
- e.setNumber(m.getNumber());
- e.setNumberProbability(m.getNumberProb());
+ if (e.getNumberProbability() < nProb) {
+ e.setNumber(n);
+ e.setNumberProbability(nProb);
}
e.addMention(m);
dm.mentionEntity(e);
} else {
- logger.debug("Creating Extent: {} {} {}", m.toText(), m.getGender(),
m.getNumber());
- e = new DiscourseEntity(m, m.getGender(), m.getGenderProb(),
m.getNumber(), m.getNumberProb());
+ logger.debug("Creating Extent: {} {} {}", m.toText(), g, n);
+ e = new DiscourseEntity(m, g, gProb, n, nProb);
dm.addEntity(e);
}
} else {
+ DiscourseEntity newEntity = new DiscourseEntity(m, g, gProb, n, nProb);
+ dm.addEntity(newEntity);
if (e != null) {
- DiscourseEntity newEntity =
- new DiscourseEntity(m, m.getGender(), m.getGenderProb(),
m.getNumber(), m.getNumberProb());
- dm.addEntity(newEntity);
newEntity.setId(e.getId());
}
- else {
- DiscourseEntity newEntity =
- new DiscourseEntity(m, m.getGender(), m.getGenderProb(),
m.getNumber(), m.getNumberProb());
- dm.addEntity(newEntity);
- }
}
}
@@ -269,10 +271,10 @@ public abstract class AbstractLinker implements Linker {
contexts[mi].setId(mentions[mi].getId());
mentionInSentenceIndex++;
if (mode != LinkerMode.SIM) {
- Gender g = computeGender(contexts[mi]);
- contexts[mi].setGender(g.getType(),g.getConfidence());
+ Gender g = computeGender(contexts[mi]);
+ contexts[mi].setGender(g.getType(), g.getConfidence());
Number n = computeNumber(contexts[mi]);
- contexts[mi].setNumber(n.getType(),n.getConfidence());
+ contexts[mi].setNumber(n.getType(), n.getConfidence());
}
}
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/DefaultLinker.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/DefaultLinker.java
index d113b33..83a0070 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/DefaultLinker.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/DefaultLinker.java
@@ -52,7 +52,7 @@ import org.slf4j.LoggerFactory;
* <p>
* Non-constituent entities such as pre-nominal named-entities and sub
entities in simple coordinated
* noun phases will be created. This {@link Linker} requires that named-entity
information also be provided.
- * This information can be added to the parse using the -parse option with
EnglishNameFinder.
+ * This information can be added to the parse using the {@code -parse} option
with EnglishNameFinder.
*
* @see Linker
*/
@@ -170,14 +170,14 @@ public class DefaultLinker extends AbstractLinker {
} else {
NonReferentialResolver nrr = new
FixedNonReferentialResolver(fixedNonReferentialProbability);
resolvers = new MaxentResolver[] {
- new SingularPronounResolver(corefProject, ResolverMode.TEST,nrr),
- new ProperNounResolver(corefProject, ResolverMode.TEST,nrr),
- new DefiniteNounResolver(corefProject, ResolverMode.TEST,nrr),
- new IsAResolver(corefProject, ResolverMode.TEST,nrr),
- new PluralPronounResolver(corefProject, ResolverMode.TEST,nrr),
- new PluralNounResolver(corefProject, ResolverMode.TEST,nrr),
- new CommonNounResolver(corefProject, ResolverMode.TEST,nrr),
- new SpeechPronounResolver(corefProject, ResolverMode.TEST,nrr)
+ new SingularPronounResolver(corefProject, ResolverMode.TEST, nrr),
+ new ProperNounResolver(corefProject, ResolverMode.TEST, nrr),
+ new DefiniteNounResolver(corefProject, ResolverMode.TEST, nrr),
+ new IsAResolver(corefProject, ResolverMode.TEST, nrr),
+ new PluralPronounResolver(corefProject, ResolverMode.TEST, nrr),
+ new PluralNounResolver(corefProject, ResolverMode.TEST, nrr),
+ new CommonNounResolver(corefProject, ResolverMode.TEST, nrr),
+ new SpeechPronounResolver(corefProject, ResolverMode.TEST, nrr)
};
}
if (LinkerMode.EVAL == mode) {
diff --git a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/Linker.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/Linker.java
index 973c184..ee6073d 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/Linker.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/Linker.java
@@ -66,7 +66,6 @@ public interface Linker {
*/
String PRONOUN_MODIFIER = "np";
-
/**
* Indicated that the specified mentions can be used to train this linker.
* This requires that the coreference relationship between the mentions have
been labeled
@@ -85,12 +84,12 @@ public interface Linker {
DiscourseEntity[] getEntities(Mention[] mentions);
/**
- * Creates mention contexts for the specified mention exents.
+ * Creates mention contexts for the specified mention extents.
* These are used to compute coreference features over.
*
* @param mentions The mention of a document.
*
- * @return mention contexts for the specified mention exents.
+ * @return mention contexts for the specified mention extents.
*/
MentionContext[] constructMentionContexts(Mention[] mentions);
@@ -110,8 +109,6 @@ public interface Linker {
MentionFinder getMentionFinder();
/**
- * Returns the head finder associated with this linker.
- *
* @return The head finder associated with this linker.
*/
HeadFinder getHeadFinder();
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/LinkerMode.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/LinkerMode.java
index 086404c..b392813 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/linker/LinkerMode.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/linker/LinkerMode.java
@@ -19,7 +19,7 @@
package opennlp.tools.coref.linker;
/**
- * Enumeration of modes in which a linker can run.
+ * Enumeration of modes in which a {@link Linker} can run.
*/
public enum LinkerMode {
@@ -33,7 +33,9 @@ public enum LinkerMode {
*/
TRAIN,
- /** Evaluation mode, used to evaluate identified coreference relationships
based on annotated text. */
+ /**
+ * Evaluation mode, used to evaluate identified coreference relationships
based on annotated text.
+ */
EVAL,
/**
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/HeadFinder.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/HeadFinder.java
index 5820c68..9ee9c99 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/HeadFinder.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/HeadFinder.java
@@ -19,13 +19,15 @@ package opennlp.tools.coref.mention;
/**
* Interface for finding head words in noun phrases and head noun-phrases in
parses.
+ *
+ * @see Parse
*/
public interface HeadFinder {
/**
- * Returns the child parse which contains the lexical head of the specified
parse.
+ * Returns the child parse which contains the lexical head of the specified
{@link Parse}.
*
- * @param parse The parse in which to find the head.
+ * @param parse The {@link Parse} in which to find the head.
* @return The parse containing the lexical head of the specified parse. If
no head is
* available or the constituent has no subcomponents that are eligible heads
then null is returned.
*/
@@ -40,18 +42,18 @@ public interface HeadFinder {
int getHeadIndex(Parse parse);
/**
- * Returns the parse bottom-most head of a <code>Parse</code>. If no
+ * Returns the parse bottom-most head of a {@link Parse}. If no
* head is available which is a child of <code>p</code> then <code>p</code>
is returned.
*
- * @param p Parse to find the head of.
+ * @param p The {@link Parse} to find the head of.
* @return bottom-most head of p.
*/
Parse getLastHead(Parse p);
/**
- * Returns head token for the specified np parse.
+ * Returns head token for the specified np {@link Parse}.
*
- * @param np The noun parse to get head from.
+ * @param np The noun {@link Parse} to get head from.
* @return head token parse.
*/
Parse getHeadToken(Parse np);
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Mention.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Mention.java
index 1101c54..c41cf2e 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Mention.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Mention.java
@@ -36,8 +36,8 @@ public class Mention implements Comparable<Mention> {
protected final String type;
/**
- * The entity id indicating which entity this extent belongs to. This is
only
- * used when training a coreference classifier.
+ * The entity id indicating which entity this extent belongs to.
+ * This is only used when training a coreference classifier.
*/
private int id;
@@ -126,7 +126,6 @@ public class Mention implements Comparable<Mention> {
return id;
}
-
@Override
public int compareTo(Mention e) {
return span.compareTo(e.span);
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionContext.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionContext.java
index d8db7ee..69c8f97 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionContext.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionContext.java
@@ -19,7 +19,6 @@ package opennlp.tools.coref.mention;
import java.util.List;
-import opennlp.tools.coref.resolver.AbstractResolver;
import opennlp.tools.coref.sim.Context;
import opennlp.tools.coref.sim.GenderEnum;
import opennlp.tools.coref.sim.NumberEnum;
@@ -36,7 +35,8 @@ public class MentionContext extends Context {
private static final Logger logger =
LoggerFactory.getLogger(MentionContext.class);
/**
- * The index of first token which is not part of a descriptor. This is 0 if
no descriptor is present.
+ * The index of first token which is not part of a descriptor.
+ * This is 0 if no descriptor is present.
*/
private final int nonDescriptorStart;
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionFinder.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionFinder.java
index 047f258..5e158db 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionFinder.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/MentionFinder.java
@@ -20,46 +20,49 @@ package opennlp.tools.coref.mention;
/**
* Specifies the interface that Objects which determine the space of
- * mentions for coreference should implement.
+ * {@link Mention mentions} for coreference should implement.
+ *
+ * @see Mention
*/
public interface MentionFinder {
/**
- * Specifies whether pre-nominal named-entities should be collected as
mentions.
+ * Specifies whether pre-nominal named-entities should be collected as
{@link Mention mentions}.
*
- * @param collectPrenominalNamedEntities true if pre-nominal named-entities
- * should be collected; false
otherwise.
+ * @param collectPrenominalNamedEntities {@code true} if pre-nominal
named-entities
+ * should be collected; {@code false}
otherwise.
*/
void setPrenominalNamedEntityCollection(boolean
collectPrenominalNamedEntities);
/**
- * Returns whether this mention finder collects pre-nominal named-entities
as mentions.
+ * Returns whether this mention finder collects pre-nominal named-entities
as {@link Mention mentions}.
*
- * @return true if this mention finder collects pre-nominal named-entities
as mentions
+ * @return {@code true} if this mention finder collects pre-nominal
named-entities as mentions.
*/
boolean isPrenominalNamedEntityCollection();
/**
- * Returns whether this mention finder collects coordinated noun phrases as
mentions.
+ * Returns whether this mention finder collects coordinated noun phrases as
{@link Mention mentions}.
*
- * @return true if this mention finder collects coordinated noun phrases as
mentions; false otherwise.
+ * @return {@code true} if this mention finder collects coordinated noun
phrases as mentions;
+ * {@code false} otherwise.
*/
boolean isCoordinatedNounPhraseCollection();
/**
- * Specifies whether coordinated noun phrases should be collected as
mentions.
+ * Specifies whether coordinated noun phrases should be collected as {@link
Mention mentions}.
*
- * @param collectCoordinatedNounPhrases true if coordinated noun phrases
should be collected;
- * false otherwise.
+ * @param collectCoordinatedNounPhrases {@code true} if coordinated noun
phrases should be collected;
+ * {@code false} otherwise.
*/
void setCoordinatedNounPhraseCollection(boolean
collectCoordinatedNounPhrases);
/**
- * Returns an array of mentions.
+ * Returns an array of {@link Mention mentions}.
*
* @param parse A top level parse from which mentions are gathered.
*
- * @return an array of mentions which implement the <code>Extent</code>
interface.
+ * @return An array of {@link Mention mentions} which implement the
<code>Extent</code> interface.
*/
Mention[] getMentions(Parse parse);
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/PTBMentionFinder.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/PTBMentionFinder.java
index 4793024..29a18d8 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/PTBMentionFinder.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/PTBMentionFinder.java
@@ -18,7 +18,10 @@
package opennlp.tools.coref.mention;
/**
- * Finds mentions from Penn Treebank style parses.
+ * A {@link MentionFinder} implementation that finds mentions from Penn
Treebank style parses.
+ *
+ * @see AbstractMentionFinder
+ * @see MentionFinder
*/
public class PTBMentionFinder extends AbstractMentionFinder {
@@ -35,10 +38,10 @@ public class PTBMentionFinder extends AbstractMentionFinder
{
}
/**
- * Retrives the one and only existing instance.
+ * Retrieves the one and only existing instance.
*
- * @param hf
- * @return the one and only existing instance
+ * @param hf A valid {@link HeadFinder} to assign if no instance exists.
+ * @return The one and only existing {@link PTBMentionFinder} instance.
*/
public static PTBMentionFinder getInstance(HeadFinder hf) {
if (instance == null) {
@@ -50,9 +53,6 @@ public class PTBMentionFinder extends AbstractMentionFinder {
return instance;
}
-
-
-
/*
private boolean isTraceNp(Parse np){
List sc = np.getSyntacticChildren();
diff --git a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Parse.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Parse.java
index c0b6c81..67f8d03 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Parse.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/Parse.java
@@ -109,28 +109,28 @@ public interface Parse extends Comparable<Parse> {
/**
* Specifies whether this parse is a named-entity.
*
- * @return {@code True} if this parse is a named-entity; {@code false}
otherwise.
+ * @return {@code true} if this parse is a named-entity; {@code false}
otherwise.
*/
boolean isNamedEntity();
/**
* Specifies whether this parse is a noun phrase.
*
- * @return {@code True} if this parse is a noun phrase; {@code false}
otherwise.
+ * @return {@code true} if this parse is a noun phrase; {@code false}
otherwise.
*/
boolean isNounPhrase();
/**
* Specifies whether this parse is a sentence.
*
- * @return {@code True} if this parse is a sentence; {@code false} otherwise.
+ * @return {@code true} if this parse is a sentence; {@code false} otherwise.
*/
boolean isSentence();
/**
* Specifies whether this parse is a coordinated noun phrase.
*
- * @return {@code True} if this parse is a coordinated noun phrase; {@code
false} otherwise.
+ * @return {@code true} if this parse is a coordinated noun phrase; {@code
false} otherwise.
*/
boolean isCoordinatedNounPhrase();
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/ShallowParseMentionFinder.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/ShallowParseMentionFinder.java
index eff4068..e652eda 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/mention/ShallowParseMentionFinder.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/mention/ShallowParseMentionFinder.java
@@ -18,7 +18,10 @@
package opennlp.tools.coref.mention;
/**
- * Finds mentions from shallow np-chunking based parses.
+ * A {@link MentionFinder} implementation that finds mentions from shallow
np-chunking based parses.
+ *
+ * @see AbstractMentionFinder
+ * @see MentionFinder
*/
public class ShallowParseMentionFinder extends AbstractMentionFinder {
@@ -32,7 +35,7 @@ public class ShallowParseMentionFinder extends
AbstractMentionFinder {
/**
* @param hf A valid {@link HeadFinder} to assign if no instance exists.
- * @return Retrieves the one and only existing instance.
+ * @return Retrieves the one and only existing {@link
ShallowParseMentionFinder} instance.
*/
public static ShallowParseMentionFinder getInstance(HeadFinder hf) {
if (instance == null) {
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/AbstractResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/AbstractResolver.java
index 197570f..d0c84c3 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/AbstractResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/AbstractResolver.java
@@ -165,6 +165,7 @@ public abstract class AbstractResolver implements Resolver {
mention.getIndexSpan().getEnd() <= cec.getIndexSpan().getEnd();
}
+ @Override
public DiscourseEntity retain(MentionContext mention, DiscourseModel dm) {
int ei = 0;
if (mention.getId() == -1) {
@@ -200,6 +201,6 @@ public abstract class AbstractResolver implements Resolver {
return fs.toString();
}
-
+ @Override
public void train() throws IOException {}
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/CommonNounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/CommonNounResolver.java
index c8f5a05..00e8357 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/CommonNounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/CommonNounResolver.java
@@ -28,18 +28,21 @@ import opennlp.tools.coref.mention.MentionContext;
* Resolves coreference between common nouns.
*
* @see MaxentResolver
+ * @see Resolver
*/
public class CommonNounResolver extends MaxentResolver {
+ private static final String MODEL_NAME = "cmodel";
+
public CommonNounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory,"cmodel", m, 80, true);
+ super(modelDirectory, MODEL_NAME, m, 80, true);
showExclusions = false;
preferFirstReferent = true;
}
public CommonNounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
throws IOException {
- super(modelDirectory,"cmodel", m, 80, true, nrr);
+ super(modelDirectory, MODEL_NAME, m, 80, true, nrr);
showExclusions = false;
preferFirstReferent = true;
}
@@ -54,12 +57,11 @@ public class CommonNounResolver extends MaxentResolver {
return features;
}
+ @Override
public boolean canResolve(MentionContext mention) {
String firstTok = mention.getFirstTokenText().toLowerCase();
String firstTokTag = mention.getFirstToken().getSyntacticType();
- boolean rv = mention.getHeadTokenTag().equals("NN")
- && !ResolverUtils.definiteArticle(firstTok, firstTokTag);
- return rv;
+ return mention.getHeadTokenTag().equals(NN) &&
!ResolverUtils.definiteArticle(firstTok, firstTokTag);
}
@Override
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefaultNonReferentialResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefaultNonReferentialResolver.java
index cc7df52..6b2fce6 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefaultNonReferentialResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefaultNonReferentialResolver.java
@@ -22,7 +22,6 @@ import java.io.BufferedReader;
import java.io.DataInputStream;
import java.io.File;
import java.io.FileInputStream;
-import java.io.FileReader;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
@@ -36,6 +35,9 @@ import java.util.ArrayList;
import java.util.List;
import java.util.zip.GZIPInputStream;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
import opennlp.tools.coref.mention.MentionContext;
import opennlp.tools.coref.mention.Parse;
import opennlp.tools.ml.maxent.GISModel;
@@ -47,8 +49,6 @@ import opennlp.tools.ml.model.FileEventStream;
import opennlp.tools.ml.model.MaxentModel;
import opennlp.tools.util.ObjectStreamUtils;
import opennlp.tools.util.TrainingParameters;
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
/**
* Default implementation of the {@link NonReferentialResolver} interface.
@@ -97,7 +97,7 @@ public class DefaultNonReferentialResolver implements
NonReferentialResolver {
List<String> features = getFeatures(mention);
double r = model.eval(features.toArray(new String[0]))[nonRefIndex];
if (DEBUG) {
- logger.debug(this + " {} -> null {} {}", mention.toText(), r, features);
+ logger.debug("{} {} -> null {} {}", this, mention.toText(), r,
features);
}
return r;
}
@@ -121,9 +121,9 @@ public class DefaultNonReferentialResolver implements
NonReferentialResolver {
}
/**
- * Returns a list of features used to predict whether the specified mention
is non-referential.
* @param mention The mention under consideration.
- * @return a list of features used to predict whether the specified mention
is non-referential.
+ * @return Retrieves a list of features used to predict whether the
+ * specified mention is non-referential.
*/
protected List<String> getNonReferentialFeatures(MentionContext mention) {
List<String> features = new ArrayList<>();
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefiniteNounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefiniteNounResolver.java
index 683e0c1..25a86c2 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefiniteNounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/DefiniteNounResolver.java
@@ -31,14 +31,16 @@ import opennlp.tools.coref.mention.MentionContext;
*/
public class DefiniteNounResolver extends MaxentResolver {
+ private static final String MODEL_NAME = "defmodel";
+
public DefiniteNounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory, "defmodel", m, 80);
+ super(modelDirectory, MODEL_NAME, m, 80);
// preferFirstReferent = true;
}
public DefiniteNounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
throws IOException {
- super(modelDirectory, "defmodel", m, 80,nrr);
+ super(modelDirectory, MODEL_NAME, m, 80,nrr);
// preferFirstReferent = true;
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/IsAResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/IsAResolver.java
index 4406d1e..5ba4a5d 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/IsAResolver.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/IsAResolver.java
@@ -55,7 +55,7 @@ public class IsAResolver extends MaxentResolver {
@Override
public boolean canResolve(MentionContext ec) {
- if (ec.getHeadTokenTag().startsWith("NN")) {
+ if (ec.getHeadTokenTag().startsWith(NN)) {
return ec.getPreviousToken() != null
&&
predicativePattern.matcher(ec.getPreviousToken().toString()).matches();
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/MaxentResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/MaxentResolver.java
index 6dd2d30..20086ca 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/MaxentResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/MaxentResolver.java
@@ -106,8 +106,11 @@ public abstract class MaxentResolver extends
AbstractResolver {
* Creates a maximum-entropy-based resolver which will look the specified
number of
* entities back for a referent. This constructor is only used for unit
testing.
*
- * @param numberOfEntitiesBack
- * @param preferFirstReferent
+ * @param numberOfEntitiesBack The number of sentences back this resolver
should look for a referent.
+ * @param preferFirstReferent If {@code true}, this designates that the
resolver should use the first
+ * referent encountered which it more preferable
than non-reference.
+ * When {@code false} all non-excluded referents
within this resolvers
+ * range are considered.
*/
protected MaxentResolver(int numberOfEntitiesBack, boolean
preferFirstReferent) {
super(numberOfEntitiesBack);
@@ -163,7 +166,8 @@ public abstract class MaxentResolver extends
AbstractResolver {
* @param numberEntitiesBack The number of entities back in the text that
this resolver will look for a referent.
* @throws IOException If the model file is not found or can not be written
to.
*/
- public MaxentResolver(String modelDirectory, String modelName, ResolverMode
mode, int numberEntitiesBack) throws IOException {
+ public MaxentResolver(String modelDirectory, String modelName, ResolverMode
mode, int numberEntitiesBack)
+ throws IOException {
this(modelDirectory, modelName, mode, numberEntitiesBack, false);
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/NonReferentialResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/NonReferentialResolver.java
index d365a85..1a7b910 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/NonReferentialResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/NonReferentialResolver.java
@@ -23,9 +23,11 @@ import opennlp.tools.coref.mention.MentionContext;
/**
* Provides the interface for an object to provide a resolver with a
non-referential
- * probability. Non-referential resolvers compute the probability that a
particular mention refers
- * to no antecedent. This probability can then compete with the probability
that
+ * probability. Non-referential resolvers compute the probability that a
particular mention refers
+ * to no antecedent. This probability can then compete with the probability
that
* a mention refers with a specific antecedent.
+ *
+ * @see MentionContext
*/
public interface NonReferentialResolver {
@@ -40,13 +42,13 @@ public interface NonReferentialResolver {
/**
* Designates that the specified mention be used for training.
*
- * @param mention The mention to be used. The mention id is used to
determine
+ * @param mention The mention to be used. The mention id is used to determine
* whether this mention is referential or non-referential.
*/
void addEvent(MentionContext mention);
/**
- * Trains a model based on the events given to this resolver via #addEvent.
+ * Trains a model based on the events given to this resolver via {@link
#addEvent}.
*
* @throws IOException When the model can not be written out.
*/
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PerfectResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PerfectResolver.java
index 5d3053d..28d32b2 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PerfectResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PerfectResolver.java
@@ -23,22 +23,22 @@ import opennlp.tools.coref.mention.MentionContext;
/**
* Resolver used in training to update the discourse model based on the
coreference annotation.
+ *
+ * @see AbstractResolver
+ * @see Resolver
*/
-public class PerfectResolver extends AbstractResolver {
+public class PerfectResolver extends AbstractResolver {
public PerfectResolver() {
super(0);
}
+ @Override
public boolean canResolve(MentionContext ec) {
return true;
}
@Override
- protected boolean outOfRange(MentionContext ec, DiscourseEntity de) {
- return false;
- }
-
public DiscourseEntity resolve(MentionContext ec, DiscourseModel dm) {
return null;
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralNounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralNounResolver.java
index ae815ea..f7d7aee 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralNounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralNounResolver.java
@@ -28,17 +28,20 @@ import opennlp.tools.coref.mention.MentionContext;
* Resolves coreference between plural nouns.
*
* @see MaxentResolver
+ * @see Resolver
*/
public class PluralNounResolver extends MaxentResolver {
+ private static final String MODEL_NAME = "plmodel";
+
public PluralNounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory, "plmodel", m, 80, true);
+ super(modelDirectory, MODEL_NAME, m, 80, true);
showExclusions = false;
}
public PluralNounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
throws IOException {
- super(modelDirectory, "plmodel", m, 80, true, nrr);
+ super(modelDirectory, MODEL_NAME, m, 80, true, nrr);
showExclusions = false;
}
@@ -58,8 +61,7 @@ public class PluralNounResolver extends MaxentResolver {
public boolean canResolve(MentionContext mention) {
String firstTok = mention.getFirstTokenText().toLowerCase();
String firstTokTag = mention.getFirstToken().getSyntacticType();
- return mention.getHeadTokenTag().equals("NNS")
- && !ResolverUtils.definiteArticle(firstTok, firstTokTag);
+ return NNS.equals(mention.getHeadTokenTag()) &&
!ResolverUtils.definiteArticle(firstTok, firstTokTag);
}
@Override
@@ -69,7 +71,7 @@ public class PluralNounResolver extends MaxentResolver {
}
else {
MentionContext cec = entity.getLastExtent();
- return (!cec.getHeadTokenTag().equals("NNS") || super.excluded(mention,
entity));
+ return (!NNS.equals(cec.getHeadTokenTag()) || super.excluded(mention,
entity));
}
}
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralPronounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralPronounResolver.java
index a72ea5f..b7a6c17 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralPronounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/PluralPronounResolver.java
@@ -28,18 +28,20 @@ import opennlp.tools.coref.mention.MentionContext;
* Resolves coreference between plural pronouns and their referents.
*
* @see MaxentResolver
+ * @see Resolver
*/
public class PluralPronounResolver extends MaxentResolver {
+ private static final String MODEL_NAME = "tmodel";
private static final int NUM_SENTS_BACK_PRONOUNS = 2;
public PluralPronounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory, "tmodel", m, 30);
+ super(modelDirectory, MODEL_NAME, m, 30);
}
public PluralPronounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
throws IOException {
- super(modelDirectory, "tmodel", m, 30, nrr);
+ super(modelDirectory, MODEL_NAME, m, 30, nrr);
}
@Override
@@ -89,7 +91,7 @@ public class PluralPronounResolver extends MaxentResolver {
@Override
public boolean canResolve(MentionContext mention) {
String tag = mention.getHeadTokenTag();
- return tag != null && tag.startsWith("PRP")
+ return tag != null && tag.startsWith(PRP)
&&
ResolverUtils.PLURAL_THIRD_PERSON_PRONOUN_PATTERN.matcher(mention.getHeadTokenText()).matches();
}
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ProperNounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ProperNounResolver.java
index 3c96b08..7fbf558 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ProperNounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ProperNounResolver.java
@@ -40,16 +40,18 @@ import org.slf4j.LoggerFactory;
* Resolves coreference between proper nouns.
*
* @see MaxentResolver
+ * @see Resolver
*/
public class ProperNounResolver extends MaxentResolver {
private static final Logger logger =
LoggerFactory.getLogger(ProperNounResolver.class);
+ private static final String MODEL_NAME = "pnmodel";
private static Map<String, Set<String>> acroMap;
private static boolean acroMapLoaded = false;
public ProperNounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory,"pnmodel", m, 500);
+ super(modelDirectory, MODEL_NAME, m, 500);
if (!acroMapLoaded) {
initAcronyms(modelDirectory + "/acronyms");
acroMapLoaded = true;
@@ -59,7 +61,7 @@ public class ProperNounResolver extends MaxentResolver {
public ProperNounResolver(String modelDirectory, ResolverMode
m,NonReferentialResolver nonRefResolver)
throws IOException {
- super(modelDirectory,"pnmodel", m, 500,nonRefResolver);
+ super(modelDirectory, MODEL_NAME, m, 500,nonRefResolver);
if (!acroMapLoaded) {
initAcronyms(modelDirectory + "/acronyms");
acroMapLoaded = true;
@@ -69,7 +71,7 @@ public class ProperNounResolver extends MaxentResolver {
@Override
public boolean canResolve(MentionContext mention) {
- return (mention.getHeadTokenTag().startsWith("NNP") ||
mention.getHeadTokenTag().startsWith("CD"));
+ return (mention.getHeadTokenTag().startsWith(NNP) ||
mention.getHeadTokenTag().startsWith("CD"));
}
private void initAcronyms(String name) {
@@ -137,7 +139,7 @@ public class ProperNounResolver extends MaxentResolver {
for (Iterator<MentionContext> ei = entity.getMentions(); ei.hasNext();) {
MentionContext xec = ei.next();
- if (xec.getHeadTokenTag().startsWith("NNP")) {
+ if (xec.getHeadTokenTag().startsWith(NNP)) {
// || initialCaps.matcher(xec.headToken.toString()).find()) {
logger.debug("Kept {} with {}", xec.toText(), xec.getHeadTokenTag());
return false;
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/Resolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/Resolver.java
index 62f0b89..7bc7129 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/Resolver.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/Resolver.java
@@ -28,6 +28,11 @@ import opennlp.tools.coref.mention.MentionContext;
*/
public interface Resolver {
+ String NN = "NN"; // tag for singular noun
+ String NNP = "NNP"; // tag for proper noun
+ String NNS = "NNS"; // tag for plural nouns
+ String PRP = "PRP"; // tag for pronoun
+
/**
* Returns {@code true} if this resolver is able to resolve the referring
expression of the same type
* as the specified mention.
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ResolverUtils.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ResolverUtils.java
index aa22f79..5bed71b 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ResolverUtils.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/ResolverUtils.java
@@ -373,7 +373,7 @@ public class ResolverUtils {
String type;
for (int i = start; i < end;i++) {
type = mtokens[start].getSyntacticType();
- if (type.startsWith("NNP")) {
+ if (type.startsWith(Resolver.NNP)) {
break;
}
start++;
@@ -412,7 +412,7 @@ public class ResolverUtils {
for (Iterator<MentionContext> ei = de.getMentions(); ei.hasNext();) {
MentionContext xec = ei.next();
String xecHeadTag = xec.getHeadTokenTag();
- if (xecHeadTag.startsWith("NNP") ||
INITIAL_CAPS.matcher(xec.getHeadTokenText()).find()) {
+ if (xecHeadTag.startsWith(Resolver.NNP) ||
INITIAL_CAPS.matcher(xec.getHeadTokenText()).find()) {
return xec;
}
}
@@ -461,12 +461,12 @@ public class ResolverUtils {
public static List<String> getPronounMatchFeatures(MentionContext mention,
DiscourseEntity entity) {
boolean foundCompatiblePronoun = false;
boolean foundIncompatiblePronoun = false;
- if (mention.getHeadTokenTag().startsWith("PRP")) {
+ if (mention.getHeadTokenTag().startsWith(Resolver.PRP)) {
Map<String, String> pronounMap =
getPronounFeatureMap(mention.getHeadTokenText());
logger.debug("PronounMap: {}", pronounMap);
for (Iterator<MentionContext> mi = entity.getMentions();mi.hasNext();) {
MentionContext candidateMention = mi.next();
- if (candidateMention.getHeadTokenTag().startsWith("PRP")) {
+ if (candidateMention.getHeadTokenTag().startsWith(Resolver.PRP)) {
if
(mention.getHeadTokenText().equalsIgnoreCase(candidateMention.getHeadTokenText()))
{
foundCompatiblePronoun = true;
break;
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SingularPronounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SingularPronounResolver.java
index 0a1b635..f338f7b 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SingularPronounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SingularPronounResolver.java
@@ -32,19 +32,21 @@ import org.slf4j.LoggerFactory;
* This class resolver singular pronouns such as "he", "she", "it" and their
various forms.
*
* @see MaxentResolver
+ * @see Resolver
*/
public class SingularPronounResolver extends MaxentResolver {
private static final Logger logger =
LoggerFactory.getLogger(SingularPronounResolver.class);
+ private static final String MODEL_NAME = "pmodel";
public SingularPronounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory, "pmodel", m, 30);
+ super(modelDirectory, MODEL_NAME, m, 30);
this.numSentencesBack = 2;
}
- public SingularPronounResolver(String modelDirectory, ResolverMode m,
- NonReferentialResolver
nonReferentialResolver) throws IOException {
- super(modelDirectory, "pmodel", m, 30,nonReferentialResolver);
+ public SingularPronounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
+ throws IOException {
+ super(modelDirectory, MODEL_NAME, m, 30, nrr);
this.numSentencesBack = 2;
}
@@ -52,7 +54,7 @@ public class SingularPronounResolver extends MaxentResolver {
public boolean canResolve(MentionContext mention) {
logger.debug("CanResolve: ec=({}) {}", mention.getId(), mention.toText());
String tag = mention.getHeadTokenTag();
- return tag != null && tag.startsWith("PRP")
+ return tag != null && tag.startsWith(PRP)
&&
ResolverUtils.SINGULAR_THIRD_PERSON_PRONOUN_PATTERN.matcher(mention.getHeadTokenText()).matches();
}
@@ -113,7 +115,7 @@ public class SingularPronounResolver extends MaxentResolver
{
for (Iterator<MentionContext> ei = entity.getMentions(); ei.hasNext();) {
MentionContext entityMention = ei.next();
String tag = entityMention.getHeadTokenTag();
- if (tag != null && tag.startsWith("PRP")
+ if (tag != null && tag.startsWith(PRP)
&&
ResolverUtils.SINGULAR_THIRD_PERSON_PRONOUN_PATTERN.matcher(mention.getHeadTokenText()).matches())
{
if (mentionGender == null) { //lazy initialization
mentionGender =
ResolverUtils.getPronounGender(mention.getHeadTokenText());
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SpeechPronounResolver.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SpeechPronounResolver.java
index 030b803..2bb5d08 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SpeechPronounResolver.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/coref/resolver/SpeechPronounResolver.java
@@ -31,13 +31,15 @@ import org.slf4j.LoggerFactory;
* Resolves pronouns specific to quoted speech such as "you", "me", and "I".
*
* @see MaxentResolver
+ * @see Resolver
*/
public class SpeechPronounResolver extends MaxentResolver {
private static final Logger logger =
LoggerFactory.getLogger(SpeechPronounResolver.class);
+ private static final String MODEL_NAME = "fmodel";
public SpeechPronounResolver(String modelDirectory, ResolverMode m) throws
IOException {
- super(modelDirectory, "fmodel", m, 30);
+ super(modelDirectory, MODEL_NAME, m, 30);
this.numSentencesBack = 0;
showExclusions = false;
preferFirstReferent = true;
@@ -45,7 +47,7 @@ public class SpeechPronounResolver extends MaxentResolver {
public SpeechPronounResolver(String modelDirectory, ResolverMode m,
NonReferentialResolver nrr)
throws IOException {
- super(modelDirectory, "fmodel", m, 30, nrr);
+ super(modelDirectory, MODEL_NAME, m, 30, nrr);
showExclusions = false;
preferFirstReferent = true;
}
@@ -57,10 +59,10 @@ public class SpeechPronounResolver extends MaxentResolver {
features.addAll(ResolverUtils.getPronounMatchFeatures(mention,entity));
List<String> contexts = ResolverUtils.getContextFeatures(mention);
MentionContext cec = entity.getLastExtent();
- if (mention.getHeadTokenTag().startsWith("PRP") &&
cec.getHeadTokenTag().startsWith("PRP")) {
+ if (mention.getHeadTokenTag().startsWith(PRP) &&
cec.getHeadTokenTag().startsWith(PRP)) {
features.add(mention.getHeadTokenText() + "," +
cec.getHeadTokenText());
}
- else if (mention.getHeadTokenText().startsWith("NNP")) {
+ else if (mention.getHeadTokenText().startsWith(NNP)) {
features.addAll(contexts);
features.add(mention.getNameType() + "," + cec.getHeadTokenText());
}
@@ -82,9 +84,9 @@ public class SpeechPronounResolver extends MaxentResolver {
@Override
public boolean canResolve(MentionContext mention) {
String tag = mention.getHeadTokenTag();
- boolean fpp = tag != null && tag.startsWith("PRP")
+ boolean fpp = tag != null && tag.startsWith(PRP)
&&
ResolverUtils.SPEECH_PRONOUN_PATTERN.matcher(mention.getHeadTokenText()).matches();
- boolean pn = tag != null && tag.startsWith("NNP");
+ boolean pn = tag != null && tag.startsWith(NNP);
return (fpp || pn);
}
@@ -97,8 +99,8 @@ public class SpeechPronounResolver extends MaxentResolver {
if (!canResolve(cec)) {
return true;
}
- if (mention.getHeadTokenTag().startsWith("NNP")) { //mention is a
propernoun
- if (cec.getHeadTokenTag().startsWith("NNP")) {
+ if (mention.getHeadTokenTag().startsWith(NNP)) { //mention is a proper noun
+ if (cec.getHeadTokenTag().startsWith(NNP)) {
return true; // both NNP
}
else {
@@ -108,13 +110,13 @@ public class SpeechPronounResolver extends MaxentResolver
{
return !canResolve(cec);
}
}
- else if (mention.getHeadTokenTag().startsWith("PRP")) { // mention is a
speech pronoun
+ else if (mention.getHeadTokenTag().startsWith(PRP)) { // mention is a
speech pronoun
// cec can be either a speech pronoun or a proper noun
- if (cec.getHeadTokenTag().startsWith("NNP")) {
+ if (cec.getHeadTokenTag().startsWith(NNP)) {
//exclude antecedents not in the same sentence when they are not
pronoun
return (mention.getSentenceNumber() - cec.getSentenceNumber() != 0);
}
- else if (cec.getHeadTokenTag().startsWith("PRP")) {
+ else if (cec.getHeadTokenTag().startsWith(PRP)) {
return false;
}
else {
diff --git a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Context.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Context.java
index fc130c7..9bcc234 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Context.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Context.java
@@ -28,6 +28,7 @@ import opennlp.tools.coref.dictionary.DictionaryFactory;
import opennlp.tools.coref.mention.HeadFinder;
import opennlp.tools.coref.mention.Mention;
import opennlp.tools.coref.mention.Parse;
+import opennlp.tools.coref.resolver.Resolver;
import opennlp.tools.util.Span;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@@ -77,7 +78,7 @@ public class Context extends Mention {
tokens = tokenList.toArray(new Parse[0]);
this.headTokenTag = headToken.getSyntacticType();
this.headTokenText = headToken.toString();
- if (headTokenTag.startsWith("NN") && !headTokenTag.startsWith("NNP")) {
+ if (headTokenTag.startsWith(Resolver.NN) &&
!headTokenTag.startsWith(Resolver.NNP)) {
this.synsets = getSynsetSet(this);
}
else {
@@ -138,10 +139,10 @@ public class Context extends Mention {
Dictionary dict = DictionaryFactory.getDictionary();
logger.debug("{} lemmas for {}", lemmas.length, c.getHeadTokenText());
for (String lemma : lemmas) {
- String senseKey = dict.getSenseKey(lemma, "NN", 0);
+ String senseKey = dict.getSenseKey(lemma, Resolver.NN, 0);
if (senseKey != null) {
synsetSet.add(senseKey);
- String[] synsets = dict.getParentSenseKeys(lemma, "NN", 0);
+ String[] synsets = dict.getParentSenseKeys(lemma, Resolver.NN, 0);
synsetSet.addAll(Arrays.asList(synsets));
}
}
@@ -150,7 +151,7 @@ public class Context extends Mention {
private static String[] getLemmas(Context c) {
String word = c.headTokenText.toLowerCase();
- return DictionaryFactory.getDictionary().getLemmas(word,"NN");
+ return DictionaryFactory.getDictionary().getLemmas(word, Resolver.NN);
}
/**
diff --git a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Gender.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Gender.java
index cdff989..0a04bf5 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Gender.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Gender.java
@@ -19,6 +19,8 @@ package opennlp.tools.coref.sim;
/**
* Class which models the gender of an entity and the confidence of that
association.
+ *
+ * @see GenderEnum
*/
public class Gender {
diff --git a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Number.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Number.java
index 68a9e20..4d8e47c 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Number.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/Number.java
@@ -18,6 +18,8 @@
package opennlp.tools.coref.sim;
/**
* Class which models the number of an entity and the confidence of that
association.
+ *
+ * @see NumberEnum
*/
public class Number {
private final NumberEnum type;
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/TrainModel.java
b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/TrainModel.java
index a9b6136..0061a8a 100644
--- a/opennlp-coref/src/main/java/opennlp/tools/coref/sim/TrainModel.java
+++ b/opennlp-coref/src/main/java/opennlp/tools/coref/sim/TrainModel.java
@@ -35,7 +35,7 @@ public interface TrainModel<T>{
* <p>
* Incompatible extents are chosen at random from the set of extents which
don't meet these criteria.
*
- * @param extents
+ * @param extents The {@link Context extents} to set and process.
*/
void setExtents(Context[] extents);
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/formats/muc/MucCorefContentHandler.java
b/opennlp-coref/src/main/java/opennlp/tools/formats/muc/MucCorefContentHandler.java
index baed3f6..063ffee 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/formats/muc/MucCorefContentHandler.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/formats/muc/MucCorefContentHandler.java
@@ -31,6 +31,10 @@ import opennlp.tools.util.Span;
// Take care for special @ sign handling (identifies a table or something else
that should be ignored)
class MucCorefContentHandler extends SgmlParser.ContentHandler {
+ private static final String ID = "ID";
+ private static final String REF = "REF";
+ private static final String MIN = "MIN";
+
public static class CorefMention {
Span span;
int id;
@@ -58,11 +62,11 @@ class MucCorefContentHandler extends
SgmlParser.ContentHandler {
private RawCorefSample sample;
/**
- * Initializes a {@link MucCorefContentHandler}.
+ * Initializes a {@link MucCorefContentHandler} with the specified
parameters.
*
* @param tokenizer The {@link Tokenizer} to use. Must not be {@code null}.
- * @param samples The {@link List< RawCorefSample > samples} as input.
- * Must not be {@code null}.
+ * @param samples The {@link RawCorefSample samples} as input.
+ * Must not be {@code null}.
*/
MucCorefContentHandler(Tokenizer tokenizer, List<RawCorefSample> samples) {
this.tokenizer = tokenizer;
@@ -72,7 +76,7 @@ class MucCorefContentHandler extends
SgmlParser.ContentHandler {
/**
* Resolves an id via the references to the root {@code id}.
*
- * @param id the id or reference to be resolved
+ * @param id the id or reference to be resolved.
*
* @return the resolved {@code id} or {@code -1} if id cannot be resolved.
*/
@@ -104,8 +108,8 @@ class MucCorefContentHandler extends
SgmlParser.ContentHandler {
if (COREF_ELEMENT.equals(name)) {
int beginOffset = text.size();
- String idString = attributes.get("ID");
- String refString = attributes.get("REF");
+ String idString = attributes.get(ID);
+ String refString = attributes.get(REF);
int id;
if (idString != null) {
@@ -124,7 +128,7 @@ class MucCorefContentHandler extends
SgmlParser.ContentHandler {
// throw invalid format exception ...
}
- mentionStack.push(new CorefMention(new Span(beginOffset, beginOffset),
id, attributes.get("MIN")));
+ mentionStack.push(new CorefMention(new Span(beginOffset, beginOffset),
id, attributes.get(MIN)));
}
}
diff --git
a/opennlp-coref/src/main/java/opennlp/tools/lang/english/TreebankNameFinder.java
b/opennlp-coref/src/main/java/opennlp/tools/lang/english/TreebankNameFinder.java
index 7d4739a..ff9fb28 100644
---
a/opennlp-coref/src/main/java/opennlp/tools/lang/english/TreebankNameFinder.java
+++
b/opennlp-coref/src/main/java/opennlp/tools/lang/english/TreebankNameFinder.java
@@ -23,15 +23,15 @@ import java.io.IOException;
import java.io.InputStreamReader;
import java.nio.file.FileSystems;
-import opennlp.tools.coref.sim.SimilarityModel;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
import opennlp.tools.namefind.NameFinderEventStream;
import opennlp.tools.namefind.NameFinderME;
import opennlp.tools.namefind.TokenNameFinderModel;
import opennlp.tools.parser.Parse;
import opennlp.tools.tokenize.SimpleTokenizer;
import opennlp.tools.util.Span;
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
/**
* Class is used to create a name finder for English.
@@ -42,6 +42,8 @@ import org.slf4j.LoggerFactory;
public class TreebankNameFinder {
private static final Logger logger =
LoggerFactory.getLogger(TreebankNameFinder.class);
+ private static final String SYMBOL_LT = "<";
+ private static final String SYMBOL_GT = ">";
public static String[] NAME_TYPES =
{"person", "organization", "location", "date", "time", "percentage",
"money"};
@@ -90,11 +92,13 @@ public class TreebankNameFinder {
}
/**
- * Adds sgml style name tags to the specified input buffer and outputs this
information to stdout.
+ * Adds sgml style name {@code tags} to the specified {@code input} buffer
and outputs this information to stdout.
+ *
* @param finders The name finders to be used.
* @param tags The tag names for the corresponding name finder.
- * @param input The input reader.
- * @throws IOException
+ * @param input The {@link BufferedReader input reader}.
+ *
+ * @throws IOException Thrown if IO errors occurred.
*/
private static void processText(TreebankNameFinder[] finders, String[] tags,
BufferedReader input)
throws IOException {
@@ -124,7 +128,7 @@ public class TreebankNameFinder {
|| nameOutcomes[fi][ti].equals(NameFinderME.OTHER))
&& (nameOutcomes[fi][ti - 1].equals(NameFinderME.START)
|| nameOutcomes[fi][ti - 1].equals(NameFinderME.CONTINUE))) {
- output.append("</").append(tags[fi]).append(">");
+ output.append("</").append(tags[fi]).append(SYMBOL_GT);
}
}
}
@@ -134,7 +138,7 @@ public class TreebankNameFinder {
//check for start tags
for (int fi = 0, fl = finders.length; fi < fl; fi++) {
if (nameOutcomes[fi][ti].equals(NameFinderME.START)) {
- output.append("<").append(tags[fi]).append(">");
+ output.append(SYMBOL_LT).append(tags[fi]).append(SYMBOL_GT);
}
}
output.append(tokens[ti]);
@@ -144,7 +148,7 @@ public class TreebankNameFinder {
for (int fi = 0, fl = finders.length; fi < fl; fi++) {
if (nameOutcomes[fi][tokens.length - 1].equals(NameFinderME.START)
|| nameOutcomes[fi][tokens.length -
1].equals(NameFinderME.CONTINUE)) {
- output.append("</").append(tags[fi]).append(">");
+ output.append("</").append(tags[fi]).append(SYMBOL_GT);
}
}
}
diff --git
a/opennlp-coref/src/test/java/opennlp/tools/coref/linker/AbstractLinkerTest.java
b/opennlp-coref/src/test/java/opennlp/tools/coref/linker/AbstractLinkerTest.java
index 925fc0e..824dcc8 100644
---
a/opennlp-coref/src/test/java/opennlp/tools/coref/linker/AbstractLinkerTest.java
+++
b/opennlp-coref/src/test/java/opennlp/tools/coref/linker/AbstractLinkerTest.java
@@ -54,7 +54,7 @@ public abstract class AbstractLinkerTest extends
AbstractCorefTest {
static void showEntities(DiscourseEntity[] entities) {
for (int ei = 0, en = entities.length; ei < en; ei++) {
- logger.debug(ei + " " + entities[ei]);
+ logger.debug("{} {}", ei, entities[ei]);
}
}
}
