List of usage examples for edu.stanford.nlp.ling IndexedWord ner
@Override
public String ner()
From source file:ca.ualberta.exemplar.core.RelationExtraction.java
License:Open Source License
private static boolean hasNerChild(IndexedWord word, SemanticGraph dependencies) { for (IndexedWord child : dependencies.getChildren(word)) { if (child.ner() != null) { return true; }//from w w w . ja va 2s . co m } return false; }
From source file:ca.ualberta.exemplar.core.RelationExtraction.java
License:Open Source License
private static void addModifiers(List<IndexedWord> words, IndexedWord word, SemanticGraph dependencies) { List<IndexedWord> adjs = dependencies.getChildrenWithReln(word, GrammaticalRelation.valueOf("amod")); List<IndexedWord> nns = dependencies.getChildrenWithReln(word, GrammaticalRelation.valueOf("nn")); List<IndexedWord> negs = dependencies.getChildrenWithReln(word, GrammaticalRelation.valueOf("neg")); List<IndexedWord> pvts = dependencies.getChildrenWithReln(word, GrammaticalRelation.valueOf("pvt")); // phrasal verb particle -- shut down List<IndexedWord> newWords = new ArrayList<IndexedWord>(); if (adjs != null) newWords.addAll(adjs);/*from w w w .j a v a2 s . c om*/ if (nns != null) newWords.addAll(nns); if (negs != null) newWords.addAll(negs); if (pvts != null) newWords.addAll(pvts); for (IndexedWord newWord : newWords) { if (Math.abs(word.index() - newWord.index()) > 5) { // If a modifier is too far way from trigger (> 5 tokens), ignore this modifier since it is probably a mistake continue; } if (!newWord.ner().equals("PERSON") && !newWord.ner().equals("ORGANIZATION") && !newWord.ner().equals("LOCATION") && !newWord.ner().equals("MISC")) { words.add(newWord); } } }
From source file:featureExtractor.NLPFeatures.java
static void processLine(String text, int lineId) throws IOException { bw_root.write(Integer.toString(lineId)); bw_subj.write(Integer.toString(lineId)); bw_underRoot.write(Integer.toString(lineId)); bw_nerType.write(Integer.toString(lineId)); //text = "A gigantic Hong Kong set was constructed in downtown Detroit. The set was so big that the Detroit People Mover track ended up becoming part of the set and shooting had to be adjusted to allow the track to move through the set. ";//"One of three new television series scheduled for release in 2014 based on DC Comics characters. The others being Constantine (2014) and The Flash (2014). "; HashMap<String, Integer> nerCount = new HashMap<>(); int superlativePOS = 0; try {//from ww w. j a va 2 s .com Annotation document = new Annotation(text); pipeline.annotate(document); List<CoreMap> sentences = document.get(CoreAnnotations.SentencesAnnotation.class); for (CoreMap sentence : sentences) { SemanticGraph dependencies = sentence .get(SemanticGraphCoreAnnotations.CollapsedDependenciesAnnotation.class); // getting root words for (IndexedWord rword : dependencies.getRoots()) { //System.out.println(rword.lemma()); //System.out.println(rword.ner()); if (rword.ner().equals("O")) bw_root.write("\t" + rword.ner() + ":" + rword.lemma()); //else if(rword.ner().equals("PERSON")) else bw_root.write("\t" + rword.ner() + ":" + rword.originalText()); /* else bw_root.write(" entity_" + rword.ner()); */ // under root for (IndexedWord child : dependencies.getChildren(rword)) { //System.out.println("here: " + child.originalText()); /* if(child.ner().equals("PERSON")) bw_underRoot.write(" " + child.originalText()); else*/ if (!child.ner().equals("O")) bw_underRoot.write("\t" + child.ner() + ":" + child.originalText()); } // nsubj | nsubpass words GrammaticalRelation[] subjects = { EnglishGrammaticalRelations.NOMINAL_SUBJECT, EnglishGrammaticalRelations.NOMINAL_PASSIVE_SUBJECT }; for (IndexedWord current : dependencies.descendants(rword)) for (IndexedWord nsubWord : dependencies.getChildrenWithRelns(current, Arrays.asList(subjects))) { //System.out.println("wow: " + nsubWord.originalText()); if (!nsubWord.ner().equals("O")) bw_subj.write("\t" + nsubWord.ner() + ":" + nsubWord.originalText()); else { //System.out.println(nsubWord.lemma()); bw_subj.write("\t" + nsubWord.ner() + ":" + nsubWord.lemma()); } /* else bw_subj.write(" entity_"+nsubWord.ner()); */ } } // NER Types frequency for (CoreLabel token : sentence.get(CoreAnnotations.TokensAnnotation.class)) { String pos = token.get(CoreAnnotations.PartOfSpeechAnnotation.class); String ne = token.get(CoreAnnotations.NamedEntityTagAnnotation.class); if (pos.equals("JJS") || pos.equals("RBS")) superlativePOS++; nerCount.putIfAbsent(ne, 0); nerCount.put(ne, nerCount.get(ne) + 1); } //System.out.println("dependency graph:\n" + dependencies); } } catch (Exception e) { System.out.println("IGNORED:"); } bw_nerType.write("\t" + Integer.toString(superlativePOS)); for (String ne : ners) { if (nerCount.containsKey(ne)) bw_nerType.write("\t" + nerCount.get(ne).toString()); else bw_nerType.write("\t0"); } bw_root.write("\n"); bw_underRoot.write("\n"); bw_nerType.write("\n"); bw_subj.write("\n"); if (lineId % 25 == 0) { bw_root.flush(); bw_underRoot.flush(); bw_nerType.flush(); bw_subj.flush(); } }
From source file:semRewrite.datesandnumber.DatesAndDuration.java
License:Open Source License
/** *************************************************************** *//* w w w. j ava 2s.c o m*/ public void generateDurationSumoTerms(IndexedWord tempParent, Utilities utilities, semRewrite.datesandnumber.DateInfo startDateInfo, DateInfo endDateInfo) { if (tempParent != null) { if (Utilities.VerbTags.contains(tempParent.tag())) { utilities.sumoTerms.add("StartTime(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + startDateInfo.getTimeCount() + ")"); utilities.sumoTerms.add("EndTime(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + endDateInfo.getTimeCount() + ")"); } if (Utilities.nounTags.contains(tempParent.tag())) { if (tempParent.ner().equals("PERSON")) { utilities.sumoTerms.add("BirthDate(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + startDateInfo.getTimeCount() + ")"); utilities.sumoTerms.add("DeathDate(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + endDateInfo.getTimeCount() + ")"); } else { utilities.sumoTerms.add("StartTime(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + startDateInfo.getTimeCount() + ")"); utilities.sumoTerms.add("EndTime(" + tempParent.value() + "-" + tempParent.index() + "," + "time-" + endDateInfo.getTimeCount() + ")"); } } startDateInfo.setDurationFlag(true); endDateInfo.setDurationFlag(true); } }