Examples with SemanticGraph edu.stanford.nlp.semgraph.SemanticGraph used on opensource projects

Example 66 with SemanticGraph

use of edu.stanford.nlp.semgraph.SemanticGraph in project CoreNLP by stanfordnlp.

the class UniversalEnglishGrammaticalStructure method demoteQuantificationalModifiers.

private static void demoteQuantificationalModifiers(SemanticGraph sg) {
    SemanticGraph sgCopy = sg.makeSoftCopy();
    SemgrexMatcher matcher = QUANT_MOD_3W_PATTERN.matcher(sgCopy);
    while (matcher.findNextMatchingNode()) {
        IndexedWord w1 = matcher.getNode("w1");
        IndexedWord w2 = matcher.getNode("w2");
        IndexedWord w3 = matcher.getNode("w3");
        IndexedWord gov = matcher.getNode("gov");
        demoteQmodParentHelper(sg, gov, w2);
        List<IndexedWord> otherDeps = Generics.newLinkedList();
        otherDeps.add(w1);
        otherDeps.add(w2);
        otherDeps.add(w3);
        demoteQmodMWEHelper(sg, otherDeps, gov, w2);
    }
    for (SemgrexPattern p : QUANT_MOD_2W_PATTERNS) {
        sgCopy = sg.makeSoftCopy();
        matcher = p.matcher(sgCopy);
        while (matcher.findNextMatchingNode()) {
            IndexedWord w1 = matcher.getNode("w1");
            IndexedWord w2 = matcher.getNode("w2");
            IndexedWord gov = matcher.getNode("gov");
            demoteQmodParentHelper(sg, gov, w1);
            List<IndexedWord> otherDeps = Generics.newLinkedList();
            otherDeps.add(w1);
            otherDeps.add(w2);
            demoteQmodMWEHelper(sg, otherDeps, gov, w1);
        }
    }
}

Example 67 with SemanticGraph

use of edu.stanford.nlp.semgraph.SemanticGraph in project CoreNLP by stanfordnlp.

the class UniversalEnglishGrammaticalStructure method addCaseMarkerInformation.

/**
   * Adds the case marker(s) to all nmod, acl and advcl relations that are
   * modified by one or more case markers(s).
   *
   * @param enhanceOnlyNmods If this is set to true, then prepositons will only be appended to nmod
   *                         relations (and not to acl or advcl) relations.
   *
   * @see UniversalEnglishGrammaticalStructure#addCaseMarkersToReln
   */
private static void addCaseMarkerInformation(SemanticGraph sg, boolean enhanceOnlyNmods) {
    /* Semgrexes require a graph with a root. */
    if (sg.getRoots().isEmpty())
        return;
    /* passive agent */
    SemanticGraph sgCopy = sg.makeSoftCopy();
    SemgrexMatcher matcher = PASSIVE_AGENT_PATTERN.matcher(sgCopy);
    while (matcher.find()) {
        IndexedWord caseMarker = matcher.getNode("c1");
        IndexedWord gov = matcher.getNode("gov");
        IndexedWord mod = matcher.getNode("mod");
        addPassiveAgentToReln(sg, gov, mod, caseMarker);
    }
    List<IndexedWord> oldCaseMarkers = Generics.newArrayList();
    /* 3-word prepositions */
    for (SemgrexPattern p : PREP_MW3_PATTERNS) {
        sgCopy = sg.makeSoftCopy();
        matcher = p.matcher(sgCopy);
        while (matcher.find()) {
            if (enhanceOnlyNmods && !matcher.getRelnString("reln").equals("nmod")) {
                continue;
            }
            List<IndexedWord> caseMarkers = Generics.newArrayList(3);
            caseMarkers.add(matcher.getNode("c1"));
            caseMarkers.add(matcher.getNode("c2"));
            caseMarkers.add(matcher.getNode("c3"));
            Collections.sort(caseMarkers);
            /* We only want to match every case marker once. */
            if (caseMarkers.equals(oldCaseMarkers))
                continue;
            IndexedWord gov = matcher.getNode("gov");
            IndexedWord mod = matcher.getNode("mod");
            addCaseMarkersToReln(sg, gov, mod, caseMarkers);
            oldCaseMarkers = caseMarkers;
        }
    }
    /* 2-word prepositions */
    for (SemgrexPattern p : PREP_MW2_PATTERNS) {
        sgCopy = sg.makeSoftCopy();
        matcher = p.matcher(sgCopy);
        while (matcher.find()) {
            if (enhanceOnlyNmods && !matcher.getRelnString("reln").equals("nmod")) {
                continue;
            }
            List<IndexedWord> caseMarkers = Generics.newArrayList(2);
            caseMarkers.add(matcher.getNode("c1"));
            caseMarkers.add(matcher.getNode("c2"));
            Collections.sort(caseMarkers);
            /* We only want to match every case marker once. */
            if (caseMarkers.equals(oldCaseMarkers))
                continue;
            IndexedWord gov = matcher.getNode("gov");
            IndexedWord mod = matcher.getNode("mod");
            addCaseMarkersToReln(sg, gov, mod, caseMarkers);
            oldCaseMarkers = caseMarkers;
        }
    }
    /* Single-word prepositions */
    for (SemgrexPattern p : PREP_PATTERNS) {
        sgCopy = sg.makeSoftCopy();
        matcher = p.matcher(sgCopy);
        while (matcher.find()) {
            if (enhanceOnlyNmods && !matcher.getRelnString("reln").equals("nmod")) {
                continue;
            }
            List<IndexedWord> caseMarkers = Generics.newArrayList(1);
            caseMarkers.add(matcher.getNode("c1"));
            if (caseMarkers.equals(oldCaseMarkers))
                continue;
            IndexedWord gov = matcher.getNode("gov");
            IndexedWord mod = matcher.getNode("mod");
            addCaseMarkersToReln(sg, gov, mod, caseMarkers);
            oldCaseMarkers = caseMarkers;
        }
    }
}

Example 68 with SemanticGraph

use of edu.stanford.nlp.semgraph.SemanticGraph in project CoreNLP by stanfordnlp.

the class UniversalEnglishGrammaticalStructure method addEnhancements.

@Override
protected void addEnhancements(List<TypedDependency> list, EnhancementOptions options) {
    SemanticGraph sg = new SemanticGraph(list);
    if (DEBUG) {
        printListSorted("addEnhancements: before correctDependencies()", sg.typedDependencies());
    }
    correctDependencies(sg);
    if (DEBUG) {
        printListSorted("addEnhancements: after correctDependencies()", sg.typedDependencies());
    }
    /* Turn multi-word prepositions into flat mwe. */
    if (options.processMultiWordPrepositions) {
        processMultiwordPreps(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after processMultiwordPreps()", sg.typedDependencies());
        }
    }
    /* Turn quantificational modifiers into flat mwe. */
    if (options.demoteQuantMod) {
        demoteQuantificationalModifiers(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after demoteQuantificationalModifiers()", sg.typedDependencies());
        }
    }
    /* Add copy nodes for conjoined Ps and PPs. */
    if (options.addCopyNodes) {
        expandPPConjunctions(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after expandPPConjunctions()", sg.typedDependencies());
        }
        expandPrepConjunctions(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after expandPrepConjunctions()", sg.typedDependencies());
        }
    }
    /* Add propositions to relation names. */
    if (options.enhancePrepositionalModifiers) {
        addCaseMarkerInformation(sg, options.enhanceOnlyNmods);
        if (DEBUG) {
            printListSorted("addEnhancements: after addCaseMarkerInformation()", sg.typedDependencies());
        }
    }
    /* Add coordinating conjunctions to relation names. */
    if (options.enhanceConjuncts) {
        addConjInformation(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after addConjInformation()", sg.typedDependencies());
        }
    }
    /* Add "referent" relations. */
    if (options.addReferent) {
        addRef(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after addRef()", sg.typedDependencies());
        }
        collapseReferent(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after collapseReferent()", sg.typedDependencies());
        }
    }
    /* Propagate dependents. */
    if (options.propagateDependents) {
        treatCC(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after treatCC()", sg.typedDependencies());
        }
    }
    /* Add relations between controlling subjects and controlled verbs. */
    if (options.addXSubj) {
        addExtraNSubj(sg);
        if (DEBUG) {
            printListSorted("addEnhancements: after addExtraNSubj()", sg.typedDependencies());
        }
    }
    correctSubjPass(sg);
    list.clear();
    list.addAll(sg.typedDependencies());
    Collections.sort(list);
}

Example 69 with SemanticGraph

use of edu.stanford.nlp.semgraph.SemanticGraph in project CoreNLP by stanfordnlp.

the class UniversalEnglishGrammaticalStructure method postProcessDependencies.

@Override
protected void postProcessDependencies(List<TypedDependency> list) {
    SemanticGraph sg = new SemanticGraph(list);
    postProcessDependencies(sg);
    list.clear();
    list.addAll(sg.typedDependencies());
}

Example 70 with SemanticGraph

use of edu.stanford.nlp.semgraph.SemanticGraph in project CoreNLP by stanfordnlp.

the class TSVUtils method parseSentence.

/** Create an Annotation object (with a single sentence) from the given specification */
private static Annotation parseSentence(Optional<String> docid, Optional<Integer> sentenceIndex, String gloss, Function<List<CoreLabel>, SemanticGraph> tree, Function<List<CoreLabel>, SemanticGraph> maltTree, List<String> words, List<String> lemmas, List<String> pos, List<String> ner, Optional<String> sentenceid) {
    // Error checks
    if (lemmas.size() != words.size()) {
        throw new IllegalArgumentException("Array lengths don't match: " + words.size() + " vs " + lemmas.size() + " (sentence " + sentenceid.orElse("???") + ")");
    }
    if (pos.size() != words.size()) {
        throw new IllegalArgumentException("Array lengths don't match: " + words.size() + " vs " + pos.size() + " (sentence " + sentenceid.orElse("???") + ")");
    }
    if (ner.size() != words.size()) {
        throw new IllegalArgumentException("Array lengths don't match: " + words.size() + " vs " + ner.size() + " (sentence " + sentenceid.orElse("???") + ")");
    }
    // Create structure
    List<CoreLabel> tokens = new ArrayList<>(words.size());
    int beginChar = 0;
    for (int i = 0; i < words.size(); ++i) {
        CoreLabel token = new CoreLabel(12);
        token.setWord(words.get(i));
        token.setValue(words.get(i));
        token.setBeginPosition(beginChar);
        token.setEndPosition(beginChar + words.get(i).length());
        beginChar += words.get(i).length() + 1;
        token.setLemma(lemmas.get(i));
        token.setTag(pos.get(i));
        token.setNER(ner.get(i));
        token.set(CoreAnnotations.DocIDAnnotation.class, docid.orElse("???"));
        token.set(CoreAnnotations.SentenceIndexAnnotation.class, sentenceIndex.orElse(-1));
        token.set(CoreAnnotations.IndexAnnotation.class, i + 1);
        token.set(CoreAnnotations.TokenBeginAnnotation.class, i);
        token.set(CoreAnnotations.TokenEndAnnotation.class, i + 1);
        tokens.add(token);
    }
    gloss = gloss.replace("\\n", "\n").replace("\\t", "\t");
    CoreMap sentence = new ArrayCoreMap(16);
    sentence.set(CoreAnnotations.TokensAnnotation.class, tokens);
    SemanticGraph graph = tree.apply(tokens);
    sentence.set(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation.class, graph);
    sentence.set(SemanticGraphCoreAnnotations.CollapsedDependenciesAnnotation.class, graph);
    sentence.set(SemanticGraphCoreAnnotations.CollapsedCCProcessedDependenciesAnnotation.class, graph);
    SemanticGraph maltGraph = maltTree.apply(tokens);
    sentence.set(SemanticGraphCoreAnnotations.AlternativeDependenciesAnnotation.class, maltGraph);
    sentence.set(CoreAnnotations.DocIDAnnotation.class, docid.orElse("???"));
    sentence.set(CoreAnnotations.SentenceIndexAnnotation.class, sentenceIndex.orElse(-1));
    sentence.set(CoreAnnotations.TextAnnotation.class, gloss);
    sentence.set(CoreAnnotations.TokenBeginAnnotation.class, 0);
    sentence.set(CoreAnnotations.TokenEndAnnotation.class, tokens.size());
    Annotation doc = new Annotation(gloss);
    doc.set(CoreAnnotations.TokensAnnotation.class, tokens);
    doc.set(CoreAnnotations.SentencesAnnotation.class, Collections.singletonList(sentence));
    doc.set(CoreAnnotations.DocIDAnnotation.class, docid.orElse("???"));
    doc.set(CoreAnnotations.SentenceIndexAnnotation.class, sentenceIndex.orElse(-1));
    return doc;
}