Menu
▾
▴
[DL-Learner SVN] SF.net SVN: dl-learner:[4095] trunk/components-core/src/main/java/org/ dllearner/algorithms/isle
|
From: <dfl...@us...> - 2013-09-06 13:31:47
|
Revision: 4095
http://sourceforge.net/p/dl-learner/code/4095
Author: dfleischhacker
Date: 2013-09-06 13:31:43 +0000 (Fri, 06 Sep 2013)
Log Message:
-----------
Make LinguisticUtil singleton
Modified Paths:
--------------
trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/LinguisticUtil.java
trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SemanticAnnotator.java
trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SimpleEntityCandidatesTrie.java
trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/TrieLinguisticAnnotator.java
trunk/components-core/src/main/java/org/dllearner/algorithms/isle/textretrieval/AnnotationEntityTextRetriever.java
Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/LinguisticUtil.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/LinguisticUtil.java 2013-09-06 12:48:08 UTC (rev 4094)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/LinguisticUtil.java 2013-09-06 13:31:43 UTC (rev 4095)
@@ -6,23 +6,25 @@
import org.dllearner.algorithms.isle.WordNet;
import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
/**
* Provides shortcuts to commonly used linguistic operations
* @author Daniel Fleischhacker
*/
public class LinguisticUtil {
+ private static LinguisticUtil instance;
+
private static final WordNet wn = new WordNet();
private static POS[] RELEVANT_POS = new POS[]{POS.NOUN, POS.VERB};
private static Lemmatizer lemmatizer;
- static {
- try {
- lemmatizer = new DefaultLemmatizer();
+ public static LinguisticUtil getInstance() {
+ if (instance == null) {
+ instance = new LinguisticUtil();
}
- catch (Exception e) {
- e.printStackTrace();
- }
+ return instance;
}
/**
@@ -30,7 +32,7 @@
* @param camelCase the word containing camelcase to split
* @return all words as camelcase contained in the given word
*/
- public static String[] getWordsFromCamelCase(String camelCase) {
+ public String[] getWordsFromCamelCase(String camelCase) {
ArrayList<String> resultingWords = new ArrayList<String>();
StringBuilder sb = new StringBuilder();
for (int i = 0; i < camelCase.length(); i++) {
@@ -66,7 +68,7 @@
* @param underScored word to split at underscores
* @return words contained in given word
*/
- public static String[] getWordsFromUnderscored(String underScored) {
+ public String[] getWordsFromUnderscored(String underScored) {
return underScored.split("_");
}
@@ -77,7 +79,7 @@
* @param word the word to retrieve synonyms for
* @return synonyms for the given word
*/
- public static String[] getSynonymsForWord(String word) {
+ public String[] getSynonymsForWord(String word) {
ArrayList<String> synonyms = new ArrayList<String>();
for (POS pos : RELEVANT_POS) {
@@ -94,7 +96,7 @@
* @param n the number of senses to get lemmas for
* @return synonyms for the given word
*/
- public static String[] getTopSynonymsForWord(String word, int n) {
+ public String[] getTopSynonymsForWord(String word, int n) {
ArrayList<String> synonyms = new ArrayList<String>();
for (POS pos : RELEVANT_POS) {
@@ -104,30 +106,48 @@
}
/**
- * Returns the normalized form of the given word. This method is only able to work with single words! If there is an
- * error normalizing the given word, the word itself is returned.
+ * Returns the normalized form of the given word. If the word contains spaces, each part separated by spaces is
+ * normalized independently and joined afterwards. If there is an error normalizing the given word, the word itself
+ * is returned.
*
* @param word the word to get normalized form for
* @return normalized form of the word or the word itself on an error
*/
- public static String getNormalizedForm(String word) {
- try {
- if (lemmatizer == null) {
- return word;
+ public String getNormalizedForm(String word) {
+ StringBuilder res = new StringBuilder();
+
+ boolean first = true;
+
+ ArrayList<String> singleWords = new ArrayList<String>();
+ Collections.addAll(singleWords, word.split(" "));
+
+ for (String w : singleWords) {
+ try {
+ if (first) {
+ first = false;
+ }
+ else {
+ res.append(" ");
+ }
+ if (lemmatizer == null) {
+ res.append(w);
+ }
+ else {
+ res.append(lemmatizer.lemmatize(w));
+ }
}
- return lemmatizer.lemmatize(word);
+ catch (Exception e) {
+ e.printStackTrace();
+ }
}
- catch (Exception e) {
- e.printStackTrace();
- }
- return word;
+ return res.toString();
}
public static void main(String[] args) {
- System.out.println(getNormalizedForm("going"));
- for (String s : getWordsFromCamelCase("thisIsAClassWith1Name123")) {
+ System.out.println(LinguisticUtil.getInstance().getNormalizedForm("going"));
+ for (String s : LinguisticUtil.getInstance().getWordsFromCamelCase("thisIsAClassWith1Name123")) {
System.out.println(s);
- for (String w : getSynonymsForWord(s)) {
+ for (String w : LinguisticUtil.getInstance().getSynonymsForWord(s)) {
System.out.println(" --> " + w);
}
}
Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SemanticAnnotator.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SemanticAnnotator.java 2013-09-06 12:48:08 UTC (rev 4094)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SemanticAnnotator.java 2013-09-06 13:31:43 UTC (rev 4095)
@@ -42,7 +42,10 @@
Set<SemanticAnnotation> semanticAnnotations = new HashSet<SemanticAnnotation>();
for (Annotation annotation : annotations) {
Set<Entity> candidateEntities = entityCandidateGenerator.getCandidates(annotation);
- SemanticAnnotation semanticAnnotation = wordSenseDisambiguation.disambiguate(annotation, candidateEntities);
+ if (candidateEntities == null || candidateEntities.size() == 0) {
+ continue;
+ }
+ SemanticAnnotation semanticAnnotation = wordSenseDisambiguation.disambiguate(annotation, candidateEntities);
if(semanticAnnotation != null){
semanticAnnotations.add(semanticAnnotation);
}
Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SimpleEntityCandidatesTrie.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SimpleEntityCandidatesTrie.java 2013-09-06 12:48:08 UTC (rev 4094)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/SimpleEntityCandidatesTrie.java 2013-09-06 13:31:43 UTC (rev 4095)
@@ -45,14 +45,14 @@
for (Entity entity : relevantText.keySet()) {
for (String text : relevantText.get(entity)) {
- text = StringUtils.join(LinguisticUtil.getWordsFromCamelCase(text), " ");
- text = StringUtils.join(LinguisticUtil.getWordsFromUnderscored(text), " ");
+ text = StringUtils.join(LinguisticUtil.getInstance().getWordsFromCamelCase(text), " ");
+ text = StringUtils.join(LinguisticUtil.getInstance().getWordsFromUnderscored(text), " ");
if (text.trim().isEmpty()) {
continue;
}
addEntry(text, entity);
for (String alternativeText : nameGenerator.getAlternativeText(text)) {
-// System.out.println("New alternative text for " + text + " --> " + alternativeText);
+ System.out.println("New alternative text for " + text + " --> " + alternativeText);
addEntry(alternativeText, entity);
}
// Adds also composing words, e.g. for "has child", "has" and "child" are also added
@@ -60,7 +60,7 @@
for (String subtext : text.split(" ")) {
addEntry(subtext, entity);
for (String alternativeText : nameGenerator.getAlternativeText(subtext)) {
-// System.out.println("New alternative text for " + subtext + " --> " + alternativeText);
+ System.out.println("New alternative text for " + subtext + " --> " + alternativeText);
addEntry(alternativeText, entity);
}
//System.out.println("trie.add("+subtext+","++")");
@@ -146,7 +146,7 @@
@Override
public List<String> getAlternativeText(String word) {
- return Arrays.asList(LinguisticUtil.getTopSynonymsForWord(word, maxNumberOfSenses));
+ return Arrays.asList(LinguisticUtil.getInstance().getTopSynonymsForWord(word, maxNumberOfSenses));
}
}
@@ -167,10 +167,10 @@
@Override
public List<String> getAlternativeText(String word) {
ArrayList<String> res = new ArrayList<String>();
- res.add(LinguisticUtil.getNormalizedForm(word));
+ res.add(LinguisticUtil.getInstance().getNormalizedForm(word));
- for (String w : LinguisticUtil
- .getTopSynonymsForWord(LinguisticUtil.getNormalizedForm(word), maxNumberOfSenses)) {
+ for (String w : LinguisticUtil.getInstance()
+ .getTopSynonymsForWord(LinguisticUtil.getInstance().getNormalizedForm(word), maxNumberOfSenses)) {
res.add(w.replaceAll("_", " "));
}
Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/TrieLinguisticAnnotator.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/TrieLinguisticAnnotator.java 2013-09-06 12:48:08 UTC (rev 4094)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/index/TrieLinguisticAnnotator.java 2013-09-06 13:31:43 UTC (rev 4095)
@@ -27,7 +27,7 @@
Set<Annotation> annotations = new HashSet<Annotation>();
for (int i = 0; i < content.length(); i++) {
String unparsed = content.substring(i);
- String match = candidatesTrie.getLongestMatch(unparsed);
+ String match = candidatesTrie.getLongestMatch(LinguisticUtil.getInstance().getNormalizedForm(unparsed));
if (match != null && !match.isEmpty()) {
Annotation annotation = new Annotation(document, i, match.length());
annotations.add(annotation);
Modified: trunk/components-core/src/main/java/org/dllearner/algorithms/isle/textretrieval/AnnotationEntityTextRetriever.java
===================================================================
--- trunk/components-core/src/main/java/org/dllearner/algorithms/isle/textretrieval/AnnotationEntityTextRetriever.java 2013-09-06 12:48:08 UTC (rev 4094)
+++ trunk/components-core/src/main/java/org/dllearner/algorithms/isle/textretrieval/AnnotationEntityTextRetriever.java 2013-09-06 13:31:43 UTC (rev 4095)
@@ -91,8 +91,8 @@
if(textWithWeight.isEmpty() && useShortFormFallback){
String shortForm = sfp.getShortForm(IRI.create(entity.getURI()));
- shortForm = Joiner.on(" ").join(LinguisticUtil.getWordsFromCamelCase(shortForm));
- shortForm = Joiner.on(" ").join(LinguisticUtil.getWordsFromUnderscored(shortForm)).trim();
+ shortForm = Joiner.on(" ").join(LinguisticUtil.getInstance().getWordsFromCamelCase(shortForm));
+ shortForm = Joiner.on(" ").join(LinguisticUtil.getInstance().getWordsFromUnderscored(shortForm)).trim();
textWithWeight.put(shortForm, weight);
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
