svn commit: r1645742 - /ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java

[cl...@apache.org]

Author: clin
Date: Mon Dec 15 20:39:57 2014
New Revision: 1645742

URL: http://svn.apache.org/r1645742
Log:
Add a feature extractor to extract token features for any covering UMLS entities.

Added:
    ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java   (with props)

Added: ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java?rev=1645742&view=auto
==============================================================================
--- ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java (added)
+++ ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java Mon Dec 15 20:39:57 2014
@@ -0,0 +1,194 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+package org.apache.ctakes.temporal.ae.feature;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.Map;
+
+import org.apache.ctakes.constituency.parser.util.AnnotationTreeUtils;
+import org.apache.ctakes.core.util.DocumentIDAnnotationUtil;
+import org.apache.ctakes.relationextractor.ae.features.RelationFeaturesExtractor;
+import org.apache.ctakes.typesystem.type.syntax.BaseToken;
+import org.apache.ctakes.typesystem.type.syntax.TerminalTreebankNode;
+import org.apache.ctakes.typesystem.type.syntax.TreebankNode;
+import org.apache.ctakes.typesystem.type.textsem.EventMention;
+import org.apache.ctakes.typesystem.type.textsem.IdentifiedAnnotation;
+import org.apache.uima.analysis_engine.AnalysisEngineProcessException;
+import org.apache.uima.fit.util.JCasUtil;
+import org.apache.uima.jcas.JCas;
+import org.apache.uima.jcas.tcas.Annotation;
+import org.cleartk.ml.Feature;
+import org.cleartk.ml.feature.extractor.CleartkExtractor;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.Bag;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.Covered;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.FirstCovered;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.Following;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.LastCovered;
+import org.cleartk.ml.feature.extractor.CleartkExtractor.Preceding;
+import org.cleartk.ml.feature.extractor.CombinedExtractor1;
+import org.cleartk.ml.feature.extractor.CoveredTextExtractor;
+import org.cleartk.ml.feature.extractor.DistanceExtractor;
+import org.cleartk.ml.feature.extractor.FeatureExtractor1;
+import org.cleartk.ml.feature.extractor.NamingExtractor1;
+import org.cleartk.util.ViewUriUtil;
+
+import com.google.common.collect.Multimap;
+
+/**
+ * for every UMLS entities covering the annotation, extracted related token features
+ * @author CH151862
+ *
+ */
+public class MultiTokenFeaturesExtractor implements RelationFeaturesExtractor {
+
+	private FeatureExtractor1 coveredText = new CoveredTextExtractor();
+
+	/**
+	 * First word of the mention, last word of the mention, all words of the mention as a bag, the
+	 * preceding 3 words, the following 3 words
+	 */
+	private FeatureExtractor1 tokenContext = new CleartkExtractor(
+			BaseToken.class,
+			coveredText,
+			new FirstCovered(1),
+			new LastCovered(1),
+			new Bag(new Covered()),
+			new Preceding(3),
+			new Following(3));
+
+	/**
+	 * All extractors for mention 1, with features named to distinguish them from mention 2
+	 */
+	private FeatureExtractor1 mention1FeaturesExtractor = new NamingExtractor1(
+			"mention1",
+			new CombinedExtractor1(coveredText, tokenContext));
+
+	/**
+	 * All extractors for mention 2, with features named to distinguish them from mention 1
+	 */
+	private FeatureExtractor1 mention2FeaturesExtractor = new NamingExtractor1(
+			"mention2",
+			new CombinedExtractor1(coveredText, tokenContext));
+
+	/**
+	 * First word, last word, and all words between the mentions
+	 */
+	private CleartkExtractor tokensBetween = new CleartkExtractor(
+			BaseToken.class,
+			new NamingExtractor1("BetweenMentions", coveredText),
+			new FirstCovered(1),
+			new LastCovered(1),
+			new Bag(new Covered()));
+
+	private String cachedDocID = null;
+	private Map<EventMention, Collection<EventMention>> coveringMap;
+
+	/**
+	 * Number of words between the mentions
+	 */
+	private DistanceExtractor nTokensBetween = new DistanceExtractor(null, BaseToken.class);
+
+	@Override
+	public List<Feature> extract(JCas jCas, IdentifiedAnnotation mention1, IdentifiedAnnotation mention2)
+			throws AnalysisEngineProcessException {
+
+		String docId=null;
+		try{
+			docId = ViewUriUtil.getURI(jCas).toString();// get docID
+		}catch(Exception e){
+			docId = DocumentIDAnnotationUtil.getDocumentID(jCas);
+		}
+		if(!docId.equals(cachedDocID)){
+			// rebuild event-event map
+			cachedDocID = docId;
+			coveringMap = JCasUtil.indexCovering(jCas, EventMention.class, EventMention.class);
+		}
+
+		List<Feature> features = new ArrayList<>();
+		Annotation arg1 = mention1;
+		Annotation arg2 = mention2;
+
+		List<Annotation> arg1s = new ArrayList<>();		
+		if(arg1 instanceof EventMention){
+			for (EventMention event1 : coveringMap.get(arg1)){
+				Annotation tempAnn =  getExpandedEvent(jCas, event1);
+				if(tempAnn == null) tempAnn = event1;
+				arg1s.add(tempAnn);
+			}
+			arg1 = getExpandedEvent(jCas, mention1);
+			if(arg1 == null) arg1 = mention1;			
+		}
+		arg1s.add(arg1);
+
+		List<Annotation> arg2s = new ArrayList<>();		
+		if(arg2 instanceof EventMention){
+			for (EventMention event2 : coveringMap.get(arg2)){
+				Annotation tempAnn =  getExpandedEvent(jCas, event2);
+				if(tempAnn == null) tempAnn = event2;
+				arg2s.add(tempAnn);
+			}
+			arg2 = getExpandedEvent(jCas, mention2);
+			if(arg2 == null) arg2 = mention2;			
+		}
+		arg2s.add(arg2);
+
+		for (Annotation ann1 : arg1s){
+			features.addAll(this.mention1FeaturesExtractor.extract(jCas, ann1));
+			for (Annotation ann2: arg2s){
+				features.addAll(this.tokensBetween.extractBetween(jCas, ann1, ann2));
+				features.addAll(this.nTokensBetween.extract(jCas, ann1, ann2));
+			}
+		}
+		for (Annotation ann2: arg2s){
+			features.addAll(this.mention2FeaturesExtractor.extract(jCas, ann2));
+		}
+		
+		return features;
+	}
+
+	private static TreebankNode getExpandedEvent(JCas jCas, IdentifiedAnnotation mention){
+		// since events are single words, we are at a terminal node:
+		List<TerminalTreebankNode> terms = JCasUtil.selectCovered(TerminalTreebankNode.class, mention);
+		if(terms == null || terms.size() == 0){
+			return null;
+		}
+
+		TreebankNode coveringNode = AnnotationTreeUtils.annotationNode(jCas, mention);
+		if(coveringNode == null) return terms.get(0);
+
+		String pos =terms.get(0).getNodeType(); 
+		// do not expand Verbs
+		if(pos.startsWith("V")) return coveringNode;
+
+		if(pos.startsWith("N")){
+			// get first NP node:
+			while(coveringNode != null && !coveringNode.getNodeType().equals("NP")){
+				coveringNode = coveringNode.getParent();
+			}
+		}else if(pos.startsWith("J")){
+			while(coveringNode != null && !coveringNode.getNodeType().equals("ADJP")){
+				coveringNode = coveringNode.getParent();
+			}
+		}
+		if(coveringNode == null) coveringNode = terms.get(0);
+		return coveringNode;    
+	}
+}

Propchange: ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/MultiTokenFeaturesExtractor.java
------------------------------------------------------------------------------
    svn:mime-type = text/plain