Example usage for edu.stanford.nlp.ling IndexedWord index

List of usage examples for edu.stanford.nlp.ling IndexedWord index

  1. HOME
  2. Java
  3. edu.stanford.nlp
  4. edu.stanford.nlp.ling.*
  5. IndexedWord
  6. index

Introduction

In this page you can find the example usage for edu.stanford.nlp.ling IndexedWord index.

Prototype

@Override
    public int index()

Source Link

Usage

From source file:ca.ualberta.exemplar.core.ArgumentExtraction.java

License:Open Source License

private Argument getEntityFromHead(IndexedWord head, CoreMap sentence, SemanticGraph dependencies,
        String argumentType) {/*from  w  w  w .  j  a  va  2  s.co  m*/

    int startIndex = head.index() - 1; //Changing from starting at 1 to starting at 0
    int endIndex = head.index() - 1;

    List<CoreLabel> tokens = sentence.get(TokensAnnotation.class);

    CoreLabel token = tokens.get(startIndex);
    String ne = token.get(NamedEntityTagAnnotation.class);
    StringBuilder builder = new StringBuilder();
    builder.append(token.get(TextAnnotation.class));
    int startOffset = token.beginPosition();
    int endOffset = token.endPosition();

    // Look for first token of the entity.
    for (int index = startIndex - 1; index >= 0; index--) {

        token = tokens.get(index);
        String word = token.get(TextAnnotation.class);
        if (!ne.equals(token.get(NamedEntityTagAnnotation.class)))
            break;

        startIndex--;
        builder.insert(0, word + " ");
        startOffset = token.beginPosition();
    }

    for (int index = endIndex + 1; index < tokens.size(); index++) {

        token = tokens.get(index);
        String word = token.get(TextAnnotation.class);
        if (!ne.equals(token.get(NamedEntityTagAnnotation.class)))
            break;

        endIndex++;
        builder.append(" " + word);
        endOffset = token.endPosition();
    }

    String entityName = builder.toString();
    String entityType = normalizeEntityType(ne);
    String entityId = entityName + "#" + entityType;
    Argument argument = new Argument(argumentType, entityId, entityName, entityType, startIndex, endIndex,
            startOffset, endOffset);

    return argument;
}

From source file:ca.ualberta.exemplar.core.RelationExtraction.java

License:Open Source License

private static void sortWordsByIndex(List<IndexedWord> words) {
    Collections.sort(words, new Comparator<IndexedWord>() {

        @Override//  w  ww . j a v  a  2  s  . com
        public int compare(IndexedWord a0, IndexedWord a1) {
            return a0.index() - a1.index();
        }

    });
}

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);/*  w  w w.  java  2 s .  co  m*/
    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:edu.anu.spice.SpiceParser.java

License:Open Source License

/**
 * Checks if a word has a numerical modifier, and if so adds it as an object
 * with attribute//from   w w w .j ava  2s . c  o m
 */
protected void checkForNumericAttribute(ProposedTuples tuples, SemanticGraph sg, IndexedWord word) {
    if (sg.hasChildWithReln(word, UniversalEnglishGrammaticalRelations.NUMERIC_MODIFIER)) {
        IndexedWord nummod = sg.getChildWithReln(word, UniversalEnglishGrammaticalRelations.NUMERIC_MODIFIER);
        /* Prevent things like "number 5" */
        if (nummod.index() < word.index()) {
            tuples.addTuple(word, nummod);
        }
    } else if (sg.hasChildWithReln(word, SemanticGraphEnhancer.QMOD_RELATION)) {
        IndexedWord qmod = sg.getChildWithReln(word, SemanticGraphEnhancer.QMOD_RELATION);
        tuples.addTuple(word, qmod);
    }
}

From source file:edu.jhu.hlt.concrete.stanford.PreNERCoreMapWrapper.java

License:Open Source License

private List<Dependency> makeDependencies(SemanticGraph graph) {
    List<Dependency> depList = new ArrayList<Dependency>();
    for (IndexedWord root : graph.getRoots()) {
        // this mimics CoreNLP's handling
        String rel = GrammaticalRelation.ROOT.getLongName().replaceAll("\\s+", "");
        int dep = root.index() - 1;
        Dependency depend = DependencyFactory.create(dep, rel);
        depList.add(depend);//from  w w w .  j  av a2  s. com
    }
    for (SemanticGraphEdge edge : graph.edgeListSorted()) {
        String rel = edge.getRelation().toString().replaceAll("\\s+", "");
        int gov = edge.getSource().index() - 1;
        int dep = edge.getTarget().index() - 1;
        Dependency depend = DependencyFactory.create(dep, rel, gov);
        depList.add(depend);
    }
    return depList;
}

From source file:eu.ubipol.opinionmining.nlp_engine.Sentence.java

License:Open Source License

protected Sentence(SemanticGraph dependencies, int indexStart, DatabaseAdapter adp, int beginPosition) {
    IndexedWord rootWord = dependencies.getFirstRoot();
    sentenceRoot = new Token(rootWord.originalText(), rootWord.lemma(), rootWord.tag(), null, null,
            rootWord.index() + indexStart, rootWord.beginPosition(), rootWord.endPosition(), adp,
            beginPosition);//ww w.j av  a2 s. c  om
    addChildTokens(sentenceRoot, rootWord, dependencies, indexStart, adp, beginPosition);
    sentenceRoot.transferScores();
    if (sentenceRoot.isAKeyword())
        sentenceRoot.addAspectScore(sentenceRoot.getScore(), sentenceRoot.getWeight(),
                sentenceRoot.getAspect());
    indexStart += dependencies.size();
}

From source file:eu.ubipol.opinionmining.nlp_engine.Sentence.java

License:Open Source License

private void addChildTokens(Token rootToken, IndexedWord currentRoot, SemanticGraph dependencies,
        int indexStart, DatabaseAdapter adp, int beginPosition) {
    for (IndexedWord child : dependencies.getChildren(currentRoot)) {
        Token childToken = new Token(child.originalText(), child.lemma(), child.tag(), rootToken,
                dependencies.getEdge(currentRoot, child).toString(), child.index() + indexStart,
                child.beginPosition(), child.endPosition(), adp, beginPosition);
        rootToken.addChildToken(childToken);
        addChildTokens(childToken, child, dependencies, indexStart, adp, beginPosition);
    }/*from ww w.  j a va  2s  .c o  m*/
}

From source file:main.java.spatialrelex.markup.SpatialElement.java

public static SpatialElement setSpatialElementFeatures(Doc document, SpatialElement se) {
    IndexedWord iw = document.startOffsetIndexedWord.get(se.start);
    se.lemmaText = iw.lemma();/* ww w.  j  a  v a 2s  . co m*/
    se.startToken = iw.index();
    se.endToken = iw.index();
    int i = se.start + 1;
    while (i < se.end) {
        if (!document.startOffsetIndexedWord.containsKey(i)) {
            i++;
            continue;
        }

        iw = document.startOffsetIndexedWord.get(i);
        se.endToken = iw.index();
        se.lemmaText += " " + iw.lemma();
        if (iw.tag().contains("NN")) {
            se.generalInquirerCategories = GeneralInquirer
                    .getGeneralInquirerCategories(se.generalInquirerCategories, iw.value().toLowerCase());
            se = WordNet.setWordNetSynsetsAndHypernyms(se, iw.tag(), "NN");
        } else if (iw.tag().contains("VB")) {
            se.verbNetClasses = VerbNet.getVerbNetClasses(se.verbNetClasses, iw.value().toLowerCase());
            se = WordNet.setWordNetSynsetsAndHypernyms(se, iw.tag(), "VB");
        }
        List<String> tokenSRLs = document.startOffsetSRLRoles.get(i);
        i++;

        if (tokenSRLs == null)
            continue;
        for (String tokenSRL : tokenSRLs) {
            if (se.srls.contains(tokenSRL))
                continue;
            se.srls.add(tokenSRL);
        }
    }

    return se;
}

From source file:opendial.bn.values.RelationalVal.java

License:Open Source License

public int addNode(String value) {
    CoreLabel label = new CoreLabel();
    label.setWord(value);/*from w  ww. j  av  a 2s  .  c  om*/
    label.setValue(value);
    IndexedWord fword = new IndexedWord(label);
    fword.setIndex(graph.size());
    graph.addVertex(fword);
    cachedHashCode = 0;
    return fword.index();
}

From source file:org.sam_agent.csparser.ContinuousParser.java

License:Open Source License

public String stringify(Collection<IndexedWord> roots) {
    List<String> rootTokens = new ArrayList<String>();
    for (IndexedWord root : roots) {
        rootTokens.add(String.format("\"%s-%d\"", esc(root.value()), root.index()));
    }/*from   w w  w  .  j  a  v  a2  s.  co m*/
    return "[" + String.join(",", rootTokens) + "]";
}