Examples of edu.stanford.nlp.util.CoreMap

edu.stanford.nlp.util.CoreMap

Base type for all annotatable core objects. Should usually be instantiated as {@link ArrayCoreMap}. Many common key definitions live in {@link edu.stanford.nlp.ling.CoreAnnotations}, but others may be defined elsewhere. See {@link edu.stanford.nlp.ling.CoreAnnotations} for details.

Note that implementations of this interface must take care to implement equality correctly: by default, two CoreMaps are .equal if they contain the same keys and all corresponding values are .equal. Subclasses that wish to change this behavior (such as {@link HashableCoreMap}) must make sure that all other CoreMap implementations have a special case in their .equals to use that equality definition when appropriate. Similarly, care must be taken when defining hashcodes. The default hashcode is 37 * sum of all keys' hashcodes plus the sum of all values' hashcodes. However, use of this class as HashMap keys is discouraged because the hashcode can change over time. Consider using a {@link HashableCoreMap}.
@author dramage @author rafferty

    //props.setProperty("sup.relation.model", "/home/sonalg/javanlp/tmp/roth_relation_model_pipeline.ser");
    String text = "Barack Obama, a Yale professor, is president.";
    Annotation document = new Annotation(text);
    StanfordCoreNLP pipeline = new StanfordCoreNLP(props);
    pipeline.annotate(document);
    CoreMap sentence = document.get(CoreAnnotations.SentencesAnnotation.class).get(0);
    List<RelationMention> rel = sentence.get(MachineReadingAnnotations.RelationMentionsAnnotation.class);
    assertEquals(rel.get(0).getType(),"Work_For");
//    StringWriter stringWriter = new StringWriter();
//    pipeline.prettyPrint(document, new PrintWriter(stringWriter));
//    String result = stringWriter.getBuffer().toString();
//    System.out.println(result);

View Full Code Here


    StanfordCoreNLP pipeline = new StanfordCoreNLP();
    Annotation document = new Annotation("Stanford University is located in California. It is a great university.");
    pipeline.annotate(document);


    CoreMap sentence = document.get(CoreAnnotations.SentencesAnnotation.class).get(0);
    SemanticGraph g = sentence.get(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation.class);
    processSerialization(g);


    processSerialization(sentence.get(TreeCoreAnnotations.TreeAnnotation.class));
    processSerialization(sentence.get(CoreAnnotations.TokensAnnotation.class));
    processSerialization(sentence.get(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation.class));
    processSerialization(sentence);


    Object newDocument = processSerialization(document);
    assertTrue(newDocument instanceof Annotation);
    assertTrue(document.equals(newDocument));

View Full Code Here

  }


  public void testSentencesAnnotation() throws Exception {
    List<CoreLabel> words = getTestWords();


    CoreMap sentence = new ArrayCoreMap();
    sentence.set(CoreAnnotations.TokensAnnotation.class, words);
    List<CoreMap> sentences = new ArrayList<CoreMap>();
    sentences.add(sentence);
    Annotation document = new Annotation(text);
    document.set(CoreAnnotations.SentencesAnnotation.class, sentences);

View Full Code Here

      allTimeExpressions = new ArrayList<CoreMap>();
      List<CoreMap> allNumerics = new ArrayList<CoreMap>();
      for (CoreMap sentence: sentences) {
        // make sure that token character offsets align with the actual sentence text
        // They may not align due to token normalizations, such as "(" to "-LRB-".
        CoreMap alignedSentence =  NumberSequenceClassifier.alignSentence(sentence);
        // uncomment the next line for verbose dumping of tokens....
        // System.err.println("SENTENCE: " + ((ArrayCoreMap) sentence).toShorterString());
        List<CoreMap> timeExpressions =
          timexExtractor.extractTimeExpressionCoreMaps(alignedSentence, docDate, timeIndex);
        if (timeExpressions != null) {
          allTimeExpressions.addAll(timeExpressions);
          sentence.set(TimeAnnotations.TimexAnnotations.class, timeExpressions);
          for (CoreMap timeExpression:timeExpressions) {
            timeExpression.set(CoreAnnotations.SentenceIndexAnnotation.class, sentence.get(CoreAnnotations.SentenceIndexAnnotation.class));
          }
        }
        List<CoreMap> numbers = alignedSentence.get(CoreAnnotations.NumerizedTokensAnnotation.class);
        if(numbers != null){
          sentence.set(CoreAnnotations.NumerizedTokensAnnotation.class, numbers);
          allNumerics.addAll(numbers);
        }
      }

View Full Code Here

  /**
   * Helper method for people not working from a complete Annotation.
   * @return a list of CoreMap.  Each CoreMap represents a detected temporal expression.
   */
  public List<CoreMap> annotateSingleSentence(CoreMap sentence, String docDate, SUTime.TimeIndex timeIndex) {
    CoreMap annotationCopy = NumberSequenceClassifier.alignSentence(sentence);
    if (docDate.equals("")) {
      docDate = null;
    }
    return timexExtractor.extractTimeExpressionCoreMaps(annotationCopy, docDate, timeIndex);
  }

View Full Code Here

    return parse(tokens, null);
  }


  protected Tree parse(List<CoreLabel> tokens,
                       List<ParserConstraint> constraints) {
    CoreMap sent = new Annotation("");
    sent.set(CoreAnnotations.TokensAnnotation.class, tokens);
    sent.set(ParserAnnotations.ConstraintAnnotation.class, constraints);
    Annotation doc = new Annotation("");
    List<CoreMap> sents = new ArrayList<CoreMap>();
    sents.add(sent);
    doc.set(CoreAnnotations.SentencesAnnotation.class, sents);
    getParser().annotate(doc);

View Full Code Here


  public void testNoPOSParserAnnotator() throws Exception {
    Annotation document = new Annotation("John Bauer works at Stanford.");
    noPOSPipeline.annotate(document);
    assertEquals(1, document.get(CoreAnnotations.SentencesAnnotation.class).size());
    CoreMap sentence = document.get(CoreAnnotations.SentencesAnnotation.class).get(0);
    Tree parse = sentence.get(TreeCoreAnnotations.TreeAnnotation.class);
    assertEquals("(ROOT (S (NP (NNP John) (NNP Bauer)) (VP (VBZ works) (PP (IN at) (NP (NNP Stanford)))) (. .)))", parse.toString());
    List<CoreLabel> tokens = sentence.get(CoreAnnotations.TokensAnnotation.class);
    List<Label> leaves = parse.yield();
    assertEquals(6, tokens.size());
    assertEquals(6, leaves.size());
    String[] expectedTags = {"NNP", "NNP", "VBZ", "IN", "NNP", "."};
    for (int i = 0; i < tokens.size(); ++i) {

View Full Code Here

   */
  public void testConstraints() {
    String expectedResult = "(ROOT (S (NP (PRP$ My) (NN dog)) (ADVP (RB also)) (VP (VBZ likes) (S (VP (VBG eating) (NP (NN sausage))))) (. .)))";
    Annotation annotation = new Annotation("My dog also likes eating sausage.");
    noParserPipeline.annotate(annotation);
    CoreMap sentence = annotation.get(CoreAnnotations.SentencesAnnotation.class).get(0);


    parserOnlyPipeline.annotate(annotation);
    assertEquals(expectedResult, sentence.get(TreeCoreAnnotations.TreeAnnotation.class).toString());


    ParserConstraint constraint = new ParserConstraint(0, 2, "SBAR|SBAR[^a-zA-Z].*");
    List<ParserConstraint> constraints = new ArrayList<ParserConstraint>();
    constraints.add(constraint);
    sentence.set(ConstraintAnnotation.class, constraints);


    parserOnlyPipeline.annotate(annotation);
    String result = sentence.get(TreeCoreAnnotations.TreeAnnotation.class).toString();
    assertFalse("Tree should not match the original tree any more",
                expectedResult.equals(result));
    assertTrue("Tree should be forced to contain SBAR",
               result.indexOf("SBAR") >= 0);
  }

View Full Code Here

    pipeline.addAnnotator(new WordsToSentencesAnnotator(false));
    pipeline.addAnnotator(parser);
    Annotation document = new Annotation("John Bauer works at Stanford.");
    pipeline.annotate(document);
    assertEquals(1, document.get(CoreAnnotations.SentencesAnnotation.class).size());
    CoreMap sentence = document.get(CoreAnnotations.SentencesAnnotation.class).get(0);
    Tree parse = sentence.get(TreeCoreAnnotations.TreeAnnotation.class);
    assertEquals("(ROOT (S (NP (NNP John) (NNP Bauer)) (VP (VBZ works) (PP (IN at) (NP (NNP Stanford)))) (. .)))", parse.toString());
    List<CoreLabel> tokens = sentence.get(CoreAnnotations.TokensAnnotation.class);
    List<Label> leaves = parse.yield();
    assertEquals(6, tokens.size());
    assertEquals(6, leaves.size());
    String[] expectedTags = {"NNP", "NNP", "VBZ", "IN", "NNP", "."};
    for (int i = 0; i < tokens.size(); ++i) {

View Full Code Here

      int paragraphUtterIndex, String nextParagraphSpeaker, int paragraphOffset, Dictionaries dict) {
    if(!speakers.containsKey(paragraphUtterIndex)) {
      if(!nextParagraphSpeaker.equals("")) {
        speakers.put(paragraphUtterIndex, nextParagraphSpeaker);
      } else {  // find the speaker of this paragraph (John, nbc news)
        CoreMap lastSent = paragraph.get(paragraph.size()-1);
        String speaker = "";
        boolean hasVerb = false;
        for(int i = 0 ; i < lastSent.get(CoreAnnotations.TokensAnnotation.class).size() ; i++){
          CoreLabel w = lastSent.get(CoreAnnotations.TokensAnnotation.class).get(i);
          String pos = w.get(CoreAnnotations.PartOfSpeechAnnotation.class);
          String ner = w.get(CoreAnnotations.NamedEntityTagAnnotation.class);
          if(pos.startsWith("V")) {
            hasVerb = true;
            break;

View Full Code Here

0 1 2 3 4 5 6 7 8 9

TOP

Related Classes of edu.stanford.nlp.util.CoreMap

edu.stanford.nlp.dcoref.ACEMentionExtractor

edu.stanford.nlp.dcoref.CoNLLMentionExtractor

edu.stanford.nlp.dcoref.Document

edu.stanford.nlp.dcoref.MUCMentionExtractor

edu.stanford.nlp.dcoref.RuleBasedCorefMentionFinder

edu.stanford.nlp.dcoref.SieveCoreferenceSystem

edu.stanford.nlp.ie.crf.TestSequenceModel

edu.stanford.nlp.ie.machinereading.BasicRelationFeatureFactory

edu.stanford.nlp.ie.machinereading.domains.ace.AceReader

edu.stanford.nlp.ie.machinereading.EntityExtractorResultsPrinter

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.