You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@ctakes.apache.org by sw...@apache.org on 2013/01/25 00:26:50 UTC
svn commit: r1438262 - in
/incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk:
AssertionCleartkAnalysisEngine.java PolarityCleartkAnalysisEngine.java
SubjectCleartkAnalysisEngine.java
Author: swu
Date: Thu Jan 24 23:26:50 2013
New Revision: 1438262
URL: http://svn.apache.org/viewvc?rev=1438262&view=rev
Log:
some cosmetic changes, and adding the downsampling field. not implemented yet.
Modified:
incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/AssertionCleartkAnalysisEngine.java
incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/PolarityCleartkAnalysisEngine.java
incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/SubjectCleartkAnalysisEngine.java
Modified: incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/AssertionCleartkAnalysisEngine.java
URL: http://svn.apache.org/viewvc/incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/AssertionCleartkAnalysisEngine.java?rev=1438262&r1=1438261&r2=1438262&view=diff
==============================================================================
--- incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/AssertionCleartkAnalysisEngine.java (original)
+++ incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/AssertionCleartkAnalysisEngine.java Thu Jan 24 23:26:50 2013
@@ -24,6 +24,7 @@ import java.util.Collection;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
+import java.util.Random;
import org.apache.uima.jcas.tcas.Annotation;
@@ -101,6 +102,17 @@ public abstract class AssertionCleartkAn
defaultValue = "false")
boolean printErrors;
+ public static final String PARAM_PROBABILITY_OF_KEEPING_DEFAULT_EXAMPLE = "ProbabilityOfKeepingADefaultExample";
+
+ @ConfigurationParameter(
+ name = PARAM_PROBABILITY_OF_KEEPING_DEFAULT_EXAMPLE,
+ mandatory = false,
+ description = "probability that a default example should be retained for training")
+ protected double probabilityOfKeepingADefaultExample = 1.0;
+
+ protected Random coin = new Random(0);
+
+
public ConllDependencyNode findAnnotationHead(JCas jcas, Annotation annotation) {
for (ConllDependencyNode depNode : JCasUtil.selectCovered(jcas, ConllDependencyNode.class, annotation)) {
@@ -203,7 +215,7 @@ public abstract class AssertionCleartkAn
logger.info("processing next doc: " + documentId.getDocumentID());
} else
{
- logger.info("processing next doc (doc id is null)");
+ logger.warn("processing next doc (doc id is null)");
}
// // get gold standard relation instances during testing for error analysis
// if (! this.isTraining() && printErrors) {
@@ -217,20 +229,20 @@ public abstract class AssertionCleartkAn
// //categoryLookup = createCategoryLookup(goldView);
// }
- JCas identifiedAnnotationView, relationView;
+ JCas identifiedAnnotationView;
if (this.isTraining()) {
try {
- identifiedAnnotationView = relationView = jCas.getView(this.goldViewName);
+ identifiedAnnotationView = jCas.getView(this.goldViewName);
} catch (CASException e) {
throw new AnalysisEngineProcessException(e);
}
} else {
- identifiedAnnotationView = relationView = jCas;
+ identifiedAnnotationView = jCas;
}
- Map<IdentifiedAnnotation, Collection<Sentence>> coveringSentenceMap = JCasUtil.indexCovering(identifiedAnnotationView, IdentifiedAnnotation.class, Sentence.class);
- Map<Sentence, Collection<BaseToken>> tokensCoveredInSentenceMap = JCasUtil.indexCovered(identifiedAnnotationView, Sentence.class, BaseToken.class);
+// Map<IdentifiedAnnotation, Collection<Sentence>> coveringSentenceMap = JCasUtil.indexCovering(identifiedAnnotationView, IdentifiedAnnotation.class, Sentence.class);
+// Map<Sentence, Collection<BaseToken>> tokensCoveredInSentenceMap = JCasUtil.indexCovered(identifiedAnnotationView, Sentence.class, BaseToken.class);
List<Instance<String>> instances = new ArrayList<Instance<String>>();
// generate a list of training instances for each sentence in the document
@@ -243,7 +255,7 @@ public abstract class AssertionCleartkAn
}
if (entityMention.getPolarity() == -1)
{
- logger.info(String.format(" - identified annotation: [%d-%d] polarity %d (%s)",
+ logger.debug(String.format(" - identified annotation: [%d-%d] polarity %d (%s)",
entityMention.getBegin(),
entityMention.getEnd(),
entityMention.getPolarity(),
@@ -255,6 +267,8 @@ public abstract class AssertionCleartkAn
// instance.addAll(tokenFeatureExtractor.extract(jCas, entityMention));
// extract all features that require the token and sentence annotations
+
+ /*** Commented by SWU 01/24/13 -- doesn't seem to be used
Collection<Sentence> sentenceList = coveringSentenceMap.get(entityMention);
Sentence sentence = null;
if (sentenceList == null || sentenceList.isEmpty())
@@ -273,6 +287,7 @@ public abstract class AssertionCleartkAn
{
sentence = sentenceList.iterator().next();
}
+ */
//Sentence sentence = sentenceList.iterator().next();
/*
Modified: incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/PolarityCleartkAnalysisEngine.java
URL: http://svn.apache.org/viewvc/incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/PolarityCleartkAnalysisEngine.java?rev=1438262&r1=1438261&r2=1438262&view=diff
==============================================================================
--- incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/PolarityCleartkAnalysisEngine.java (original)
+++ incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/PolarityCleartkAnalysisEngine.java Thu Jan 24 23:26:50 2013
@@ -31,7 +31,7 @@ public class PolarityCleartkAnalysisEngi
instance.setOutcome(polarity);
if ("negated".equals(polarity))
{
- logger.info("TRAINING: " + polarity);
+ logger.debug("TRAINING: " + polarity);
}
this.dataWriter.write(instance);
} else
@@ -40,7 +40,7 @@ public class PolarityCleartkAnalysisEngi
int polarity = 1;
if (label!= null && label.equals("present"))
{
- polarity = 0;
+ polarity = 1;
} else if (label != null && label.equals("negated"))
{
polarity = -1;
@@ -48,7 +48,7 @@ public class PolarityCleartkAnalysisEngi
entityMention.setPolarity(polarity);
if ("negated".equals(label))
{
- logger.info(String.format("DECODING/EVAL: %s//%s [%d-%d] (%s)", label, polarity, entityMention.getBegin(), entityMention.getEnd(), entityMention.getClass().getName()));
+ logger.debug(String.format("DECODING/EVAL: %s//%s [%d-%d] (%s)", label, polarity, entityMention.getBegin(), entityMention.getEnd(), entityMention.getClass().getName()));
}
}
}
Modified: incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/SubjectCleartkAnalysisEngine.java
URL: http://svn.apache.org/viewvc/incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/SubjectCleartkAnalysisEngine.java?rev=1438262&r1=1438261&r2=1438262&view=diff
==============================================================================
--- incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/SubjectCleartkAnalysisEngine.java (original)
+++ incubator/ctakes/trunk/ctakes-assertion/src/main/java/org/apache/ctakes/assertion/medfacts/cleartk/SubjectCleartkAnalysisEngine.java Thu Jan 24 23:26:50 2013
@@ -82,7 +82,7 @@ public class SubjectCleartkAnalysisEngin
{
String subj = entityMention.getSubject();
instance.setOutcome(subj);
- logger.log(Level.INFO, String.format("[%s] expected: ''; actual: ''; features: %s",
+ logger.log(Level.DEBUG, String.format("[%s] expected: ''; actual: ''; features: %s",
this.getClass().getSimpleName(),
instance.toString()
//StringUtils.join(instance.getFeatures(), ", ")
@@ -92,7 +92,7 @@ public class SubjectCleartkAnalysisEngin
{
String label = this.classifier.classify(instance.getFeatures());
entityMention.setSubject(label);
- logger.log(Level.INFO, "SUBJECT is being set on an IdentifiedAnnotation: "+label+" "+entityMention.getSubject());
+ logger.log(Level.DEBUG, "SUBJECT is being set on an IdentifiedAnnotation: "+label+" "+entityMention.getSubject());
}
}