[DL-Learner SVN] SF.net SVN: dl-learner:[4127] trunk/components-core/src/main/java/org/ dllearner/algorithms/isle/wsd

[<dfl...@us...>]

Revision: 4127
          http://sourceforge.net/p/dl-learner/code/4127
Author:   dfleischhacker
Date:     2013-10-24 13:40:06 +0000 (Thu, 24 Oct 2013)
Log Message:
-----------
Fix ContextExtractors

* Use processed content instead of raw since annotations link to the former
* Fix bug occurring for tokens at index 0

Modified Paths:
--------------
    trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/SentenceBasedContextExtractor.java
    trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/WindowBasedContextExtractor.java

Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/SentenceBasedContextExtractor.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/SentenceBasedContextExtractor.java	2013-10-22 14:08:14 UTC (rev 4126)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/SentenceBasedContextExtractor.java	2013-10-24 13:40:06 UTC (rev 4127)
@@ -3,12 +3,6 @@
  */
 package org.dllearner.algorithms.isle.wsd;
 
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Properties;
-
-import org.dllearner.algorithms.isle.index.TextDocument;
-
 import edu.stanford.nlp.ling.CoreAnnotations.SentencesAnnotation;
 import edu.stanford.nlp.ling.CoreAnnotations.TextAnnotation;
 import edu.stanford.nlp.ling.CoreAnnotations.TokensAnnotation;
@@ -16,7 +10,12 @@
 import edu.stanford.nlp.pipeline.Annotation;
 import edu.stanford.nlp.pipeline.StanfordCoreNLP;
 import edu.stanford.nlp.util.CoreMap;
+import org.dllearner.algorithms.isle.index.TextDocument;
 
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Properties;
+
 /**
  * @author Lorenz Buehmann
  *
@@ -37,14 +36,14 @@
 	@Override
 	public List<String> extractContext(org.dllearner.algorithms.isle.index.Annotation annotation) {
 		//split text into sentences
-		List<CoreMap> sentences = getSentences(annotation.getReferencedDocument().getRawContent());
-		
+		List<CoreMap> sentences = getSentences(annotation.getReferencedDocument().getContent());
+
 		//find the sentence containing the token of the annotation
 		int tokenStart = annotation.getOffset();
 		int index = 0;
 		for (CoreMap sentence : sentences) {
 			String s = sentence.toString();
-			if (index < tokenStart && s.length() > tokenStart) {
+			if (index <= tokenStart && s.length() > tokenStart) {
 				List<String> context = new ArrayList<String>();
 				for (CoreLabel label : sentence.get(TokensAnnotation.class)) {
 					// this is the text of the token

Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/WindowBasedContextExtractor.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/WindowBasedContextExtractor.java	2013-10-22 14:08:14 UTC (rev 4126)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/wsd/WindowBasedContextExtractor.java	2013-10-24 13:40:06 UTC (rev 4127)
@@ -3,12 +3,6 @@
  */
 package org.dllearner.algorithms.isle.wsd;
 
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Properties;
-
-import org.dllearner.algorithms.isle.index.TextDocument;
-
 import edu.stanford.nlp.ling.CoreAnnotations.SentencesAnnotation;
 import edu.stanford.nlp.ling.CoreAnnotations.TextAnnotation;
 import edu.stanford.nlp.ling.CoreAnnotations.TokensAnnotation;
@@ -16,7 +10,12 @@
 import edu.stanford.nlp.pipeline.Annotation;
 import edu.stanford.nlp.pipeline.StanfordCoreNLP;
 import edu.stanford.nlp.util.CoreMap;
+import org.dllearner.algorithms.isle.index.TextDocument;
 
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Properties;
+
 /**
  * @author Lorenz Buehmann
  *
@@ -43,14 +42,14 @@
 	@Override
 	public List<String> extractContext(org.dllearner.algorithms.isle.index.Annotation annotation) {
 		// split text into sentences
-		List<CoreMap> sentences = getSentences(annotation.getReferencedDocument().getRawContent());
+		List<CoreMap> sentences = getSentences(annotation.getReferencedDocument().getContent());
 
 		// find the sentence containing the token of the annotation
 		int tokenStart = annotation.getOffset();
 		int index = 0;
 		for (CoreMap sentence : sentences) {
 			String s = sentence.toString();
-			if (index < tokenStart && s.length() > tokenStart) {
+			if (index <= tokenStart && s.length() > tokenStart) {
 				List<String> context = new ArrayList<String>();
 				for (CoreLabel label : sentence.get(TokensAnnotation.class)) {
 					// this is the text of the token
@@ -62,8 +61,8 @@
 			}
 			index += s.length();
 		}
-		throw new RuntimeException("Token " + annotation.getToken() + " not found in text "
-				+ annotation.getReferencedDocument().getRawContent());
+		throw new RuntimeException("Token " + annotation + " not found in text "
+				+ annotation.getReferencedDocument().getContent());
 
 	}
 	

This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.