Examples with IntTuple edu.stanford.nlp.util.IntTuple used on opensource projects

Example 6 with IntTuple

use of edu.stanford.nlp.util.IntTuple in project CoreNLP by stanfordnlp.

the class DocumentPreprocessor method fillMentionInfo.

/** initialize several variables for mentions
   * @throws Exception
   */
private static void fillMentionInfo(Document doc, Dictionaries dict, LogisticClassifier<String, String> singletonPredictor, HeadFinder headFinder) throws Exception {
    List<CoreMap> sentences = doc.annotation.get(SentencesAnnotation.class);
    for (int i = 0; i < doc.predictedMentions.size(); i++) {
        CoreMap sentence = sentences.get(i);
        for (int j = 0; j < doc.predictedMentions.get(i).size(); j++) {
            Mention m = doc.predictedMentions.get(i).get(j);
            // mentionsByID
            doc.predictedMentionsByID.put(m.mentionID, m);
            IntTuple pos = new IntTuple(2);
            pos.set(0, i);
            pos.set(1, j);
            // positions
            doc.positions.put(m, pos);
            // sentNum
            m.sentNum = i;
            IntTuple headPosition = new IntTuple(2);
            headPosition.set(0, i);
            headPosition.set(1, m.headIndex);
            // headPositions
            doc.mentionheadPositions.put(headPosition, m);
            m.contextParseTree = sentence.get(TreeAnnotation.class);
            //        m.sentenceWords = sentence.get(TokensAnnotation.class);
            m.basicDependency = sentence.get(BasicDependenciesAnnotation.class);
            m.enhancedDependency = sentence.get(SemanticGraphCoreAnnotations.EnhancedDependenciesAnnotation.class);
            if (m.enhancedDependency == null) {
                m.enhancedDependency = sentence.get(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation.class);
            }
            // mentionSubTree (highest NP that has the same head) if constituency tree available
            if (m.contextParseTree != null) {
                Tree headTree = m.contextParseTree.getLeaves().get(m.headIndex);
                if (headTree == null) {
                    throw new RuntimeException("Missing head tree for a mention!");
                }
                Tree t = headTree;
                while ((t = t.parent(m.contextParseTree)) != null) {
                    if (t.headTerminal(headFinder) == headTree && t.value().equals("NP")) {
                        m.mentionSubTree = t;
                    } else if (m.mentionSubTree != null) {
                        break;
                    }
                }
                if (m.mentionSubTree == null) {
                    m.mentionSubTree = headTree;
                }
            }
            m.process(dict, null, singletonPredictor);
        }
    }
    boolean hasGold = (doc.goldMentions != null);
    if (hasGold) {
        doc.goldMentionsByID = Generics.newHashMap();
        int sentNum = 0;
        for (List<Mention> golds : doc.goldMentions) {
            for (Mention g : golds) {
                doc.goldMentionsByID.put(g.mentionID, g);
                g.sentNum = sentNum;
            }
            sentNum++;
        }
    }
}

Example 7 with IntTuple

use of edu.stanford.nlp.util.IntTuple in project CoreNLP by stanfordnlp.

the class DocumentPreprocessor method findSubject.

private static boolean findSubject(Document doc, SemanticGraph dependency, IndexedWord w, int sentNum, int utterNum) {
    for (Pair<GrammaticalRelation, IndexedWord> child : dependency.childPairs(w)) {
        if (child.first().getShortName().equals("nsubj")) {
            String subjectString = child.second().word();
            // start from 1
            int subjectIndex = child.second().index();
            IntTuple headPosition = new IntTuple(2);
            headPosition.set(0, sentNum);
            headPosition.set(1, subjectIndex - 1);
            String speaker;
            if (doc.mentionheadPositions.containsKey(headPosition)) {
                speaker = Integer.toString(doc.mentionheadPositions.get(headPosition).mentionID);
            } else {
                speaker = subjectString;
            }
            doc.speakers.put(utterNum, speaker);
            return true;
        }
    }
    return false;
}

Example 8 with IntTuple

use of edu.stanford.nlp.util.IntTuple in project CoreNLP by stanfordnlp.

the class Document method extractGoldLinks.

/** Extract gold coref link information */
protected void extractGoldLinks() {
    //    List<List<Mention>> orderedMentionsBySentence = this.getOrderedMentions();
    List<Pair<IntTuple, IntTuple>> links = new ArrayList<>();
    // position of each mention in the input matrix, by id
    Map<Integer, IntTuple> positions = Generics.newHashMap();
    // positions of antecedents
    Map<Integer, List<IntTuple>> antecedents = Generics.newHashMap();
    for (int i = 0; i < goldMentions.size(); i++) {
        for (int j = 0; j < goldMentions.get(i).size(); j++) {
            Mention m = goldMentions.get(i).get(j);
            int id = m.mentionID;
            IntTuple pos = new IntTuple(2);
            pos.set(0, i);
            pos.set(1, j);
            positions.put(id, pos);
            antecedents.put(id, new ArrayList<>());
        }
    }
    //    SieveCoreferenceSystem.debugPrintMentions(System.err, "", goldOrderedMentionsBySentence);
    for (List<Mention> mentions : goldMentions) {
        for (Mention m : mentions) {
            int id = m.mentionID;
            IntTuple src = positions.get(id);
            assert (src != null);
            if (m.originalRef >= 0) {
                IntTuple dst = positions.get(m.originalRef);
                if (dst == null) {
                    throw new RuntimeException("Cannot find gold mention with ID=" + m.originalRef);
                }
                // to deal with cataphoric annotation
                while (dst.get(0) > src.get(0) || (dst.get(0) == src.get(0) && dst.get(1) > src.get(1))) {
                    Mention dstMention = goldMentions.get(dst.get(0)).get(dst.get(1));
                    m.originalRef = dstMention.originalRef;
                    dstMention.originalRef = id;
                    if (m.originalRef < 0)
                        break;
                    dst = positions.get(m.originalRef);
                }
                if (m.originalRef < 0)
                    continue;
                // A B C: if A<-B, A<-C => make a link B<-C
                for (int k = dst.get(0); k <= src.get(0); k++) {
                    for (int l = 0; l < goldMentions.get(k).size(); l++) {
                        if (k == dst.get(0) && l < dst.get(1))
                            continue;
                        if (k == src.get(0) && l > src.get(1))
                            break;
                        IntTuple missed = new IntTuple(2);
                        missed.set(0, k);
                        missed.set(1, l);
                        if (links.contains(new Pair<>(missed, dst))) {
                            antecedents.get(id).add(missed);
                            links.add(new Pair<>(src, missed));
                        }
                    }
                }
                links.add(new Pair<>(src, dst));
                assert (antecedents.get(id) != null);
                antecedents.get(id).add(dst);
                List<IntTuple> ants = antecedents.get(m.originalRef);
                assert (ants != null);
                for (IntTuple ant : ants) {
                    antecedents.get(id).add(ant);
                    links.add(new Pair<>(src, ant));
                }
            }
        }
    }
    goldLinks = links;
}

Example 9 with IntTuple

use of edu.stanford.nlp.util.IntTuple in project CoreNLP by stanfordnlp.

the class Document method findNextParagraphSpeaker.

private String findNextParagraphSpeaker(List<CoreMap> paragraph, int paragraphOffset, Dictionaries dict) {
    CoreMap lastSent = paragraph.get(paragraph.size() - 1);
    String speaker = "";
    for (CoreLabel w : lastSent.get(CoreAnnotations.TokensAnnotation.class)) {
        if (w.get(CoreAnnotations.LemmaAnnotation.class).equals("report") || w.get(CoreAnnotations.LemmaAnnotation.class).equals("say")) {
            String word = w.get(CoreAnnotations.TextAnnotation.class);
            SemanticGraph dependency = lastSent.get(SemanticGraphCoreAnnotations.EnhancedDependenciesAnnotation.class);
            IndexedWord t = dependency.getNodeByWordPattern(word);
            for (Pair<GrammaticalRelation, IndexedWord> child : dependency.childPairs(t)) {
                if (child.first().getShortName().equals("nsubj")) {
                    // start from 1
                    int subjectIndex = child.second().index();
                    IntTuple headPosition = new IntTuple(2);
                    headPosition.set(0, paragraph.size() - 1 + paragraphOffset);
                    headPosition.set(1, subjectIndex - 1);
                    if (mentionheadPositions.containsKey(headPosition) && mentionheadPositions.get(headPosition).nerString.startsWith("PER")) {
                        speaker = Integer.toString(mentionheadPositions.get(headPosition).mentionID);
                    }
                }
            }
        }
    }
    return speaker;
}

Example 10 with IntTuple

use of edu.stanford.nlp.util.IntTuple in project CoreNLP by stanfordnlp.

the class Document method extractGoldLinks.

/** Extract gold coref link information */
protected void extractGoldLinks() {
    //    List<List<Mention>> orderedMentionsBySentence = this.getOrderedMentions();
    List<Pair<IntTuple, IntTuple>> links = new ArrayList<>();
    // position of each mention in the input matrix, by id
    Map<Integer, IntTuple> positions = Generics.newHashMap();
    // positions of antecedents
    Map<Integer, List<IntTuple>> antecedents = Generics.newHashMap();
    for (int i = 0; i < goldOrderedMentionsBySentence.size(); i++) {
        for (int j = 0; j < goldOrderedMentionsBySentence.get(i).size(); j++) {
            Mention m = goldOrderedMentionsBySentence.get(i).get(j);
            int id = m.mentionID;
            IntTuple pos = new IntTuple(2);
            pos.set(0, i);
            pos.set(1, j);
            positions.put(id, pos);
            antecedents.put(id, new ArrayList<>());
        }
    }
    //    SieveCoreferenceSystem.debugPrintMentions(System.err, "", goldOrderedMentionsBySentence);
    for (List<Mention> mentions : goldOrderedMentionsBySentence) {
        for (Mention m : mentions) {
            int id = m.mentionID;
            IntTuple src = positions.get(id);
            assert (src != null);
            if (m.originalRef >= 0) {
                IntTuple dst = positions.get(m.originalRef);
                if (dst == null) {
                    throw new RuntimeException("Cannot find gold mention with ID=" + m.originalRef);
                }
                // to deal with cataphoric annotation
                while (dst.get(0) > src.get(0) || (dst.get(0) == src.get(0) && dst.get(1) > src.get(1))) {
                    Mention dstMention = goldOrderedMentionsBySentence.get(dst.get(0)).get(dst.get(1));
                    m.originalRef = dstMention.originalRef;
                    dstMention.originalRef = id;
                    if (m.originalRef < 0)
                        break;
                    dst = positions.get(m.originalRef);
                }
                if (m.originalRef < 0)
                    continue;
                // A B C: if A<-B, A<-C => make a link B<-C
                for (int k = dst.get(0); k <= src.get(0); k++) {
                    for (int l = 0; l < goldOrderedMentionsBySentence.get(k).size(); l++) {
                        if (k == dst.get(0) && l < dst.get(1))
                            continue;
                        if (k == src.get(0) && l > src.get(1))
                            break;
                        IntTuple missed = new IntTuple(2);
                        missed.set(0, k);
                        missed.set(1, l);
                        if (links.contains(new Pair<>(missed, dst))) {
                            antecedents.get(id).add(missed);
                            links.add(new Pair<>(src, missed));
                        }
                    }
                }
                links.add(new Pair<>(src, dst));
                assert (antecedents.get(id) != null);
                antecedents.get(id).add(dst);
                List<IntTuple> ants = antecedents.get(m.originalRef);
                assert (ants != null);
                for (IntTuple ant : ants) {
                    antecedents.get(id).add(ant);
                    links.add(new Pair<>(src, ant));
                }
            }
        }
    }
    goldLinks = links;
}