Example usage for edu.stanford.nlp.pipeline POSTaggerAnnotator POSTaggerAnnotator

List of usage examples for edu.stanford.nlp.pipeline POSTaggerAnnotator POSTaggerAnnotator

  1. HOME
  2. Java
  3. edu.stanford.nlp
  4. edu.stanford.nlp.pipeline.*
  5. POSTaggerAnnotator
  6. POSTaggerAnnotator

Introduction

In this page you can find the example usage for edu.stanford.nlp.pipeline POSTaggerAnnotator POSTaggerAnnotator.

Prototype

public POSTaggerAnnotator(String annotatorName, Properties props)

Source Link

Usage

From source file:edu.illinois.cs.cogcomp.comma.annotators.PreProcessor.java

License:Open Source License

public PreProcessor() throws Exception {
    System.out.println("initializing 1 ");
    // Initialise AnnotatorServices with default configurations
    Map<String, String> nonDefaultValues = new HashMap<>();
    if (CommaProperties.getInstance().useCurator()) {
        nonDefaultValues.put(CuratorConfigurator.RESPECT_TOKENIZATION.key, Configurator.TRUE);
        nonDefaultValues.put(CuratorConfigurator.CURATOR_FORCE_UPDATE.key, Configurator.FALSE);
        ResourceManager curatorConfig = (new CuratorConfigurator()).getConfig(nonDefaultValues);
        annotatorService = CuratorFactory.buildCuratorClient(curatorConfig);
    } else {/*from   ww  w.  j a  va  2  s  .  c o  m*/
        ResourceManager rm = new Stanford331Configurator().getDefaultConfig();
        String timePerSentence = Stanford331Configurator.STFRD_TIME_PER_SENTENCE.value;
        String maxParseSentenceLength = Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH.value;
        boolean throwExceptionOnSentenceLengthCheck = rm
                .getBoolean(Stanford331Configurator.THROW_EXCEPTION_ON_FAILED_LENGTH_CHECK.key);

        System.out.println("initializing 2 ");

        this.pos = new POSAnnotator();
        this.nerConll = NerAnnotatorManager.buildNerAnnotator(rm, ViewNames.NER_CONLL);
        this.shallowParser = new ChunkerAnnotator();

        Properties stanfordProps = new Properties();
        stanfordProps.put("annotators", "pos, parse");
        stanfordProps.put("parse.originalDependencies", true);
        stanfordProps.put("parse.maxlen", maxParseSentenceLength);
        stanfordProps.put("parse.maxtime", timePerSentence);
        // per sentence? could be per
        // document but no idea from
        // stanford javadoc
        POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
        ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
        int maxLength = Integer.parseInt(maxParseSentenceLength);
        this.parser = new StanfordParseHandler(posAnnotator, parseAnnotator, maxLength,
                throwExceptionOnSentenceLengthCheck);
    }
}

From source file:edu.illinois.cs.cogcomp.pipeline.main.PipelineFactory.java

License:Open Source License

/**
 * instantiate a set of annotators for use in an AnnotatorService object by default, will use
 * lazy initialization where possible -- change this behavior with the
 * {@link PipelineConfigurator#USE_LAZY_INITIALIZATION} property.
 * /*from  w  w  w.j av  a 2s.  co m*/
 * @param nonDefaultRm ResourceManager with all non-default values for Annotators
 * @return a Map from annotator view name to annotator
 */
private static Map<String, Annotator> buildAnnotators(ResourceManager nonDefaultRm) throws IOException {
    ResourceManager rm = new PipelineConfigurator()
            .getConfig(new Stanford331Configurator().getConfig(nonDefaultRm));
    String timePerSentence = rm.getString(Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
    String maxParseSentenceLength = rm.getString(Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
    boolean useLazyInitialization = rm.getBoolean(PipelineConfigurator.USE_LAZY_INITIALIZATION.key,
            PipelineConfigurator.TRUE);

    Map<String, Annotator> viewGenerators = new HashMap<>();

    if (rm.getBoolean(PipelineConfigurator.USE_POS)) {
        POSAnnotator pos = new POSAnnotator();
        viewGenerators.put(pos.getViewName(), pos);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_LEMMA)) {
        IllinoisLemmatizer lem = new IllinoisLemmatizer(rm);
        viewGenerators.put(lem.getViewName(), lem);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_SHALLOW_PARSE)) {
        viewGenerators.put(ViewNames.SHALLOW_PARSE, new ChunkerAnnotator());
    }
    if (rm.getBoolean(PipelineConfigurator.USE_NER_CONLL)) {
        NERAnnotator nerConll = NerAnnotatorManager.buildNerAnnotator(rm, ViewNames.NER_CONLL);
        viewGenerators.put(nerConll.getViewName(), nerConll);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_NER_ONTONOTES)) {
        NERAnnotator nerOntonotes = NerAnnotatorManager.buildNerAnnotator(rm, ViewNames.NER_ONTONOTES);
        viewGenerators.put(nerOntonotes.getViewName(), nerOntonotes);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_DEP)) {
        DepAnnotator dep = new DepAnnotator();
        viewGenerators.put(dep.getViewName(), dep);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_STANFORD_DEP)
            || rm.getBoolean(PipelineConfigurator.USE_STANFORD_PARSE)) {
        Properties stanfordProps = new Properties();
        stanfordProps.put("annotators", "pos, parse");
        stanfordProps.put("parse.originalDependencies", true);
        stanfordProps.put("parse.maxlen", maxParseSentenceLength);
        stanfordProps.put("parse.maxtime", timePerSentence); // per sentence? could be per
                                                             // document but no idea from
                                                             // stanford javadoc
        POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
        ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
        int maxLength = Integer.parseInt(maxParseSentenceLength);
        boolean throwExceptionOnSentenceLengthCheck = rm
                .getBoolean(Stanford331Configurator.THROW_EXCEPTION_ON_FAILED_LENGTH_CHECK.key);

        if (rm.getBoolean(PipelineConfigurator.USE_STANFORD_DEP)) {
            StanfordDepHandler depParser = new StanfordDepHandler(posAnnotator, parseAnnotator, maxLength,
                    throwExceptionOnSentenceLengthCheck);
            viewGenerators.put(depParser.getViewName(), depParser);
        }
        if (rm.getBoolean(PipelineConfigurator.USE_STANFORD_PARSE)) {
            StanfordParseHandler parser = new StanfordParseHandler(posAnnotator, parseAnnotator, maxLength,
                    throwExceptionOnSentenceLengthCheck);
            viewGenerators.put(parser.getViewName(), parser);
        }
    }

    if (rm.getBoolean(PipelineConfigurator.USE_SRL_VERB)) {
        Properties verbProps = new Properties();
        String verbType = SRLType.Verb.name();
        verbProps.setProperty(SrlConfigurator.SRL_TYPE.key, verbType);
        ResourceManager verbRm = new ResourceManager(verbProps);
        rm = Configurator.mergeProperties(rm, verbRm);
        try {
            SemanticRoleLabeler verbSrl = new SemanticRoleLabeler(rm, useLazyInitialization);
            viewGenerators.put(ViewNames.SRL_VERB, verbSrl);
        } catch (Exception e) {
            throw new IOException("SRL verb cannot init: " + e.getMessage());
        }
    }
    if (rm.getBoolean(PipelineConfigurator.USE_SRL_NOM)) {
        Properties nomProps = new Properties();
        String nomType = SRLType.Nom.name();
        nomProps.setProperty(SrlConfigurator.SRL_TYPE.key, nomType);
        ResourceManager nomRm = new ResourceManager(nomProps);
        rm = Configurator.mergeProperties(rm, nomRm);

        try {
            SemanticRoleLabeler nomSrl = new SemanticRoleLabeler(rm, useLazyInitialization);
            // note that you can't call nomSrl (or verbSrl).getViewName() as it may not be
            // initialized yet
            viewGenerators.put(ViewNames.SRL_NOM, nomSrl);
            // viewGenerators.put(ViewNames.SRL_NOM,new SrlHandler("NomSRL", "5.1.9", nomType,
            // ViewNames.SRL_NOM,
            // useLazyInitialization, rm));
        } catch (Exception e) {
            throw new IOException("SRL nom cannot init .." + e.getMessage());
        }
    }

    if (rm.getBoolean(PipelineConfigurator.USE_QUANTIFIER)) {
        Quantifier quantifierAnnotator = new Quantifier();
        viewGenerators.put(ViewNames.QUANTITIES, quantifierAnnotator);
    }

    if (rm.getBoolean(PipelineConfigurator.USE_TRANSLITERATION)) {
        for (Language lang : TransliterationAnnotator.supportedLanguages) {
            TransliterationAnnotator transliterationAnnotator = new TransliterationAnnotator(true, lang);
            viewGenerators.put(ViewNames.TRANSLITERATION + "_" + lang.getCode(), transliterationAnnotator);
        }
    }

    if (rm.getBoolean(PipelineConfigurator.USE_SRL_PREP)) {
        PrepSRLAnnotator prepSRLAnnotator = new PrepSRLAnnotator();
        viewGenerators.put(ViewNames.SRL_PREP, prepSRLAnnotator);
    }

    if (rm.getBoolean(PipelineConfigurator.USE_SRL_COMMA)) {
        CommaLabeler commaLabeler = new CommaLabeler();
        viewGenerators.put(ViewNames.SRL_COMMA, commaLabeler);
    }

    if (rm.getBoolean(PipelineConfigurator.USE_VERB_SENSE)) {
        VerbSenseAnnotator verbSense = new VerbSenseAnnotator();
        viewGenerators.put(ViewNames.VERB_SENSE, verbSense);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_MENTION)) {
        MentionAnnotator mentionAnnotator = new MentionAnnotator("ACE_TYPE");
        viewGenerators.put(ViewNames.MENTION, mentionAnnotator);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_RELATION)) {
        viewGenerators.put(ViewNames.RELATION, new RelationAnnotator(true));
    }
    if (rm.getBoolean(PipelineConfigurator.USE_TIMEX3)) {
        Properties rmProps = new TemporalChunkerConfigurator().getDefaultConfig().getProperties();
        TemporalChunkerAnnotator tca = new TemporalChunkerAnnotator(new ResourceManager(rmProps));
        viewGenerators.put(ViewNames.TIMEX3, tca);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_DATALESS_ESA)) {
        rm = new ESADatalessConfigurator().getConfig(nonDefaultRm);
        ESADatalessAnnotator esaDataless = new ESADatalessAnnotator(rm);
        viewGenerators.put(ViewNames.DATALESS_ESA, esaDataless);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_DATALESS_W2V)) {
        rm = new W2VDatalessConfigurator().getConfig(nonDefaultRm);
        W2VDatalessAnnotator w2vDataless = new W2VDatalessAnnotator(rm);
        viewGenerators.put(ViewNames.DATALESS_W2V, w2vDataless);
    }
    if (rm.getBoolean(PipelineConfigurator.USE_QUESTION_TYPER)) {
        QuestionTypeAnnotator questionTyper = new QuestionTypeAnnotator();
        viewGenerators.put(ViewNames.QUESTION_TYPE, questionTyper);
    }

    return viewGenerators;
}

From source file:org.cogcomp.re.ACEMentionReader.java

License:Open Source License

public ACEMentionReader(String file, String type) {
    readType = type;/*from   w  w w  .ja v  a  2s .co  m*/
    relations_mono = new ArrayList<>();
    relations_bi = new ArrayList<>();

    try {
        ACEReader reader = new ACEReaderWithTrueCaseFixer(file, new String[] { "bn", "nw" }, false);
        POSAnnotator pos_annotator = new POSAnnotator();
        ChunkerAnnotator chunker = new ChunkerAnnotator(true);
        chunker.initialize(new ChunkerConfigurator().getDefaultConfig());
        Datastore ds = new Datastore(new ResourceConfigurator().getDefaultConfig());
        File gazetteersResource = ds.getDirectory("org.cogcomp.gazetteers", "gazetteers", 1.6, false);
        Gazetteers gazetteers = GazetteersFactory.get(5,
                gazetteersResource.getPath() + File.separator + "gazetteers", true, Language.English);
        WordNetManager.loadConfigAsClasspathResource(true);
        WordNetManager wordNet = WordNetManager.getInstance();
        Properties stanfordProps = new Properties();
        stanfordProps.put("annotators", "pos, parse");
        stanfordProps.put("parse.originalDependencies", true);
        stanfordProps.put("parse.maxlen", Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
        stanfordProps.put("parse.maxtime", Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
        POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
        ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
        StanfordDepHandler stanfordDepHandler = new StanfordDepHandler(posAnnotator, parseAnnotator);

        for (TextAnnotation ta : reader) {
            if (ta.getId().equals("bn\\CNN_ENG_20030424_070008.15.apf.xml")) {
                continue;
            }
            ta.addView(pos_annotator);
            stanfordDepHandler.addView(ta);
            chunker.addView(ta);

            View entityView = ta.getView(ViewNames.MENTION_ACE);
            View annotatedTokenView = new SpanLabelView("RE_ANNOTATED", ta);
            for (Constituent co : ta.getView(ViewNames.TOKENS).getConstituents()) {
                Constituent c = co.cloneForNewView("RE_ANNOTATED");
                for (String s : co.getAttributeKeys()) {
                    c.addAttribute(s, co.getAttribute(s));
                }
                c.addAttribute("WORDNETTAG", BIOFeatureExtractor.getWordNetTags(wordNet, c));
                c.addAttribute("WORDNETHYM", BIOFeatureExtractor.getWordNetHyms(wordNet, c));
                annotatedTokenView.addConstituent(c);
            }
            ta.addView("RE_ANNOTATED", annotatedTokenView);

            List<Relation> existRelations = entityView.getRelations();
            for (int i = 0; i < ta.getNumberOfSentences(); i++) {
                Sentence curSentence = ta.getSentence(i);
                List<Constituent> cins = entityView.getConstituentsCoveringSpan(curSentence.getStartSpan(),
                        curSentence.getEndSpan());
                for (int j = 0; j < cins.size(); j++) {
                    for (int k = j + 1; k < cins.size(); k++) {
                        Constituent firstArg = cins.get(j);
                        Constituent secondArg = cins.get(k);
                        Constituent firstArgHead = RelationFeatureExtractor
                                .getEntityHeadForConstituent(firstArg, firstArg.getTextAnnotation(), "A");
                        Constituent secondArgHead = RelationFeatureExtractor
                                .getEntityHeadForConstituent(secondArg, secondArg.getTextAnnotation(), "A");
                        firstArg.addAttribute("GAZ",
                                ((FlatGazetteers) gazetteers).annotatePhrase(firstArgHead));
                        secondArg.addAttribute("GAZ",
                                ((FlatGazetteers) gazetteers).annotatePhrase(secondArgHead));

                        boolean found_as_source = false;
                        boolean found_as_target = false;
                        for (Relation r : existRelations) {
                            if (r.getSource().getStartSpan() == firstArg.getStartSpan()
                                    && r.getSource().getEndSpan() == firstArg.getEndSpan()
                                    && r.getTarget().getStartSpan() == secondArg.getStartSpan()
                                    && r.getTarget().getEndSpan() == secondArg.getEndSpan()) {
                                relations_mono.add(r);
                                found_as_source = true;
                                String opTagFine = getOppoName(r.getAttribute("RelationSubtype"));
                                String opTagCoarse = ACERelationTester.getCoarseType(opTagFine);
                                Relation opdir = new Relation(opTagCoarse, secondArg, firstArg, 1.0f);
                                opdir.addAttribute("RelationSubtype", opTagFine);
                                opdir.addAttribute("RelationType", opTagCoarse);
                                relations_bi.add(r);
                                relations_bi.add(opdir);
                                break;
                            }
                            if (r.getTarget().getStartSpan() == firstArg.getStartSpan()
                                    && r.getTarget().getEndSpan() == firstArg.getEndSpan()
                                    && r.getSource().getStartSpan() == secondArg.getStartSpan()
                                    && r.getSource().getEndSpan() == secondArg.getEndSpan()) {
                                relations_mono.add(r);
                                found_as_target = true;
                                String opTagFine = getOppoName(r.getAttribute("RelationSubtype"));
                                String opTagCoarse = ACERelationTester.getCoarseType(opTagFine);
                                Relation opdir = new Relation(opTagCoarse, firstArg, secondArg, 1.0f);
                                opdir.addAttribute("RelationSubtype", opTagFine);
                                opdir.addAttribute("RelationType", opTagCoarse);
                                relations_bi.add(r);
                                relations_bi.add(opdir);
                                break;
                            }
                        }
                        if (!found_as_source && !found_as_target) {
                            Relation newRelation_1 = new Relation("NOT_RELATED", firstArg, secondArg, 1.0f);
                            newRelation_1.addAttribute("RelationSubtype", "NOT_RELATED");
                            newRelation_1.addAttribute("RelationType", "NOT_RELATED");
                            relations_mono.add(newRelation_1);
                            Relation newRelation_2 = new Relation("NOT_RELATED", secondArg, firstArg, 1.0f);
                            newRelation_2.addAttribute("RelationSubtype", "NOT_RELATED");
                            newRelation_2.addAttribute("RelationType", "NOT_RELATED");
                            relations_bi.add(newRelation_1);
                            relations_bi.add(newRelation_2);
                        }
                    }
                }
            }
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
}

From source file:org.cogcomp.re.ExampleUsage.java

License:Open Source License

public static void AnnotatorExample() {
    String text = "He went to Chicago after his Father moved there.";

    String corpus = "story";
    String textId = "001";

    // Create a TextAnnotation From Text
    TextAnnotationBuilder stab = new TokenizerTextAnnotationBuilder(new StatefulTokenizer());
    TextAnnotation ta = stab.createTextAnnotation(corpus, textId, text);

    POSAnnotator pos_annotator = new POSAnnotator();
    ChunkerAnnotator chunker = new ChunkerAnnotator(true);
    chunker.initialize(new ChunkerConfigurator().getDefaultConfig());
    Properties stanfordProps = new Properties();
    stanfordProps.put("annotators", "pos, parse");
    stanfordProps.put("parse.originalDependencies", true);
    stanfordProps.put("parse.maxlen", Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
    stanfordProps.put("parse.maxtime", Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
    POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
    ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
    StanfordDepHandler stanfordDepHandler = new StanfordDepHandler(posAnnotator, parseAnnotator);
    RelationAnnotator relationAnnotator = new RelationAnnotator();

    try {/*from  w  w w . j  a v  a  2 s  .  co  m*/
        ta.addView(pos_annotator);
        chunker.addView(ta);
        stanfordDepHandler.addView(ta);
        relationAnnotator.addView(ta);
    } catch (Exception e) {
        e.printStackTrace();
    }

    View mentionView = ta.getView(ViewNames.MENTION);

    List<Constituent> predictedMentions = mentionView.getConstituents();
    List<Relation> predictedRelations = mentionView.getRelations();

    for (Relation r : predictedRelations) {
        IOHelper.printRelation(r);
    }
}

From source file:org.cogcomp.re.ExampleUsage.java

License:Open Source License

public static void SemEvalAnnotate() {
    String text = "People have been moving back into downtown.";
    String corpus = "semeval";
    String textId = "001";

    // Create a TextAnnotation From Text
    TextAnnotationBuilder stab = new TokenizerTextAnnotationBuilder(new StatefulTokenizer());
    TextAnnotation ta = stab.createTextAnnotation(corpus, textId, text);

    POSAnnotator pos_annotator = new POSAnnotator();
    ChunkerAnnotator chunker = new ChunkerAnnotator(true);
    chunker.initialize(new ChunkerConfigurator().getDefaultConfig());
    Properties stanfordProps = new Properties();
    stanfordProps.put("annotators", "pos, parse");
    stanfordProps.put("parse.originalDependencies", true);
    stanfordProps.put("parse.maxlen", Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
    stanfordProps.put("parse.maxtime", Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
    POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
    ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
    StanfordDepHandler stanfordDepHandler = new StanfordDepHandler(posAnnotator, parseAnnotator);
    String modelPath = "";
    FlatGazetteers gazetteers = null;/*from   w ww. java2 s .c  om*/
    try {
        ta.addView(pos_annotator);
        chunker.addView(ta);
        stanfordDepHandler.addView(ta);
        Datastore ds = new Datastore(new ResourceConfigurator().getDefaultConfig());
        File model = ds.getDirectory("org.cogcomp.re", "SEMEVAL", 1.1, false);
        modelPath = model.getPath();
        File gazetteersResource = ds.getDirectory("org.cogcomp.gazetteers", "gazetteers", 1.3, false);
        gazetteers = (FlatGazetteers) GazetteersFactory.get(5,
                gazetteersResource.getPath() + File.separator + "gazetteers", true, Language.English);
        WordNetManager.loadConfigAsClasspathResource(true);
        WordNetManager wordnet = WordNetManager.getInstance();
        View annotatedTokenView = new SpanLabelView("RE_ANNOTATED", ta);
        for (Constituent co : ta.getView(ViewNames.TOKENS).getConstituents()) {
            Constituent c = co.cloneForNewView("RE_ANNOTATED");
            for (String s : co.getAttributeKeys()) {
                c.addAttribute(s, co.getAttribute(s));
            }
            c.addAttribute("WORDNETTAG", BIOFeatureExtractor.getWordNetTags(wordnet, c));
            c.addAttribute("WORDNETHYM", BIOFeatureExtractor.getWordNetHyms(wordnet, c));
            annotatedTokenView.addConstituent(c);
        }
        ta.addView("RE_ANNOTATED", annotatedTokenView);
    } catch (Exception e) {
        e.printStackTrace();
    }

    Constituent source = new Constituent("first", "Mention", ta, 0, 1);
    Constituent target = new Constituent("second", "Mention", ta, 6, 7);
    source.addAttribute("GAZ", gazetteers.annotatePhrase(source));
    target.addAttribute("GAZ", gazetteers.annotatePhrase(target));
    Relation relation = new Relation("TEST", source, target, 1.0f);

    String prefix = modelPath + File.separator + "SEMEVAL" + File.separator + "SEMEVAL";
    semeval_relation_classifier classifier = new semeval_relation_classifier(prefix + ".lc", prefix + ".lex");
    String tag = classifier.discreteValue(relation);

    System.out.println(tag);
}

From source file:org.cogcomp.re.PredictedMentionReader.java

License:Open Source License

public PredictedMentionReader(String path) {
    relations = new ArrayList<>();
    try {//from  ww  w . j  a  v  a 2 s  . c o  m
        ACEReader aceReader = new ACEReader(path, false);
        POSAnnotator pos_annotator = new POSAnnotator();
        ChunkerAnnotator chunker = new ChunkerAnnotator(true);
        chunker.initialize(new ChunkerConfigurator().getDefaultConfig());
        MentionAnnotator mentionAnnotator = new MentionAnnotator("ACE_TYPE");

        Datastore ds = new Datastore(new ResourceConfigurator().getDefaultConfig());
        File gazetteersResource = ds.getDirectory("org.cogcomp.gazetteers", "gazetteers", 1.6, false);
        Gazetteers gazetteers = GazetteersFactory.get(5,
                gazetteersResource.getPath() + File.separator + "gazetteers", true, Language.English);
        WordNetManager.loadConfigAsClasspathResource(true);
        WordNetManager wordNet = WordNetManager.getInstance();
        Properties stanfordProps = new Properties();
        stanfordProps.put("annotators", "pos, parse");
        stanfordProps.put("parse.originalDependencies", true);
        stanfordProps.put("parse.maxlen", Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
        stanfordProps.put("parse.maxtime", Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
        POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
        ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
        StanfordDepHandler stanfordDepHandler = new StanfordDepHandler(posAnnotator, parseAnnotator);
        for (TextAnnotation ta : aceReader) {
            if (ta.getId().equals("bn\\CNN_ENG_20030424_070008.15.apf.xml")) {
                continue;
            }
            ta.addView(pos_annotator);
            mentionAnnotator.addView(ta);
            stanfordDepHandler.addView(ta);
            chunker.addView(ta);

            View annotatedTokenView = new SpanLabelView("RE_ANNOTATED", ta);
            for (Constituent co : ta.getView(ViewNames.TOKENS).getConstituents()) {
                Constituent c = co.cloneForNewView("RE_ANNOTATED");
                for (String s : co.getAttributeKeys()) {
                    c.addAttribute(s, co.getAttribute(s));
                }
                c.addAttribute("WORDNETTAG", BIOFeatureExtractor.getWordNetTags(wordNet, c));
                c.addAttribute("WORDNETHYM", BIOFeatureExtractor.getWordNetHyms(wordNet, c));
                annotatedTokenView.addConstituent(c);
            }
            ta.addView("RE_ANNOTATED", annotatedTokenView);

            View goldView = ta.getView(ViewNames.MENTION_ACE);
            View predictedView = ta.getView(ViewNames.MENTION);
            Map<Constituent, Constituent> consMap = new HashMap<Constituent, Constituent>();
            for (Constituent c : goldView.getConstituents()) {
                consMap.put(c, null);
                Constituent ch = RelationFeatureExtractor.getEntityHeadForConstituent(c, ta, "");
                for (Constituent pc : predictedView.getConstituents()) {
                    Constituent pch = MentionAnnotator.getHeadConstituent(pc, "");
                    if (ch.getStartSpan() == pch.getStartSpan() && ch.getEndSpan() == pch.getEndSpan()) {
                        consMap.put(c, pc);
                        break;
                    }
                }
            }
            size_of_gold_relations += goldView.getRelations().size();
            for (int i = 0; i < ta.getNumberOfSentences(); i++) {
                Sentence curSentence = ta.getSentence(i);
                List<Constituent> in_cur_sentence = predictedView
                        .getConstituentsCoveringSpan(curSentence.getStartSpan(), curSentence.getEndSpan());
                for (int j = 0; j < in_cur_sentence.size(); j++) {
                    for (int k = j + 1; k < in_cur_sentence.size(); k++) {
                        Constituent source = in_cur_sentence.get(j);
                        Constituent target = in_cur_sentence.get(k);
                        Constituent sourceHead = MentionAnnotator.getHeadConstituent(source, "");
                        Constituent targetHead = MentionAnnotator.getHeadConstituent(target, "");
                        source.addAttribute("GAZ", ((FlatGazetteers) gazetteers).annotatePhrase(sourceHead));
                        target.addAttribute("GAZ", ((FlatGazetteers) gazetteers).annotatePhrase(targetHead));

                        boolean found_tag = false;
                        for (Relation r : goldView.getRelations()) {

                            if (consMap.get(r.getSource()) == null || consMap.get(r.getTarget()) == null) {
                                continue;
                            }

                            Constituent gsh = ACEReader.getEntityHeadForConstituent(r.getSource(), ta, "A");
                            Constituent gth = ACEReader.getEntityHeadForConstituent(r.getTarget(), ta, "A");
                            Constituent psh = MentionAnnotator.getHeadConstituent(source, "B");
                            Constituent pth = MentionAnnotator.getHeadConstituent(target, "B");

                            if (gsh.getStartSpan() == psh.getStartSpan() && gsh.getEndSpan() == psh.getEndSpan()
                                    && gth.getStartSpan() == pth.getStartSpan()
                                    && gth.getEndSpan() == pth.getEndSpan()) {
                                Relation newRelation = new Relation(r.getAttribute("RelationSubtype"), source,
                                        target, 1.0f);
                                newRelation.addAttribute("RelationType", r.getAttribute("RelationType"));
                                newRelation.addAttribute("RelationSubtype", r.getAttribute("RelationSubtype"));
                                newRelation.addAttribute("IsGoldRelation", "True");
                                relations.add(newRelation);
                                found_tag = true;
                                break;
                            }
                            if (gsh.getStartSpan() == pth.getStartSpan() && gsh.getEndSpan() == pth.getEndSpan()
                                    && gth.getStartSpan() == psh.getStartSpan()
                                    && gth.getEndSpan() == psh.getEndSpan()) {
                                Relation newRelation = new Relation(r.getAttribute("RelationSubtype"), target,
                                        source, 1.0f);
                                newRelation.addAttribute("RelationType", r.getAttribute("RelationType"));
                                newRelation.addAttribute("RelationSubtype", r.getAttribute("RelationSubtype"));
                                newRelation.addAttribute("IsGoldRelation", "True");
                                relations.add(newRelation);
                                found_tag = true;
                                break;
                            }
                        }
                        if (!found_tag) {
                            Relation newRelation = new Relation("NOT_RELATED", source, target, 1.0f);
                            newRelation.addAttribute("RelationType", "NOT_RELATED");
                            newRelation.addAttribute("RelationSubtype", "NOT_RELATED");
                            newRelation.addAttribute("IsGoldRelation", "False");
                            relations.add(newRelation);
                        }
                    }
                }
            }
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
}

From source file:org.cogcomp.re.SemEvalMentionReader.java

License:Open Source License

public void initExternalTools() {
    try {//from w w w . j av  a2 s. c om
        _posAnnotator = new POSAnnotator();
        Datastore ds = new Datastore(new ResourceConfigurator().getDefaultConfig());
        File gazetteersResource = ds.getDirectory("org.cogcomp.gazetteers", "gazetteers", 1.3, false);
        _gazetteers = (FlatGazetteers) GazetteersFactory.get(5,
                gazetteersResource.getPath() + File.separator + "gazetteers", true, Language.English);
        WordNetManager.loadConfigAsClasspathResource(true);
        _wordnet = WordNetManager.getInstance();
        __chunker = new ChunkerAnnotator(true);
        __chunker.initialize(new ChunkerConfigurator().getDefaultConfig());

        Properties stanfordProps = new Properties();
        stanfordProps.put("annotators", "pos, parse");
        stanfordProps.put("parse.originalDependencies", true);
        stanfordProps.put("parse.maxlen", Stanford331Configurator.STFRD_MAX_SENTENCE_LENGTH);
        stanfordProps.put("parse.maxtime", Stanford331Configurator.STFRD_TIME_PER_SENTENCE);
        POSTaggerAnnotator posAnnotator = new POSTaggerAnnotator("pos", stanfordProps);
        ParserAnnotator parseAnnotator = new ParserAnnotator("parse", stanfordProps);
        __stanfordDep = new StanfordDepHandler(posAnnotator, parseAnnotator);
        __mentionAnnotator = new MentionAnnotator("ACE_TYPE");
    } catch (Exception e) {
        e.printStackTrace();
    }
}

From source file:shef.mt.tools.ParsingProcessor.java

public ParsingProcessor(String lang, String pm, String dm, HashSet<String> requirements) {
    //Store required resources:
    this.requiresPOSTags = requirements.contains("postags");
    this.requiresDepCounts = requirements.contains("depcounts");

    //Create model path objects:
    String posModel = null;/*w  ww  .  j a  va2  s. c  o  m*/
    String depModel = null;

    //Setup model paths:
    if (pm == null) {
        if (lang.equals("english")) {
            posModel = "edu/stanford/nlp/models/pos-tagger/english-left3words/english-left3words-distsim.tagger";
        } else if (lang.equals("spanish")) {
            posModel = "edu/stanford/nlp/models/pos-tagger/spanish/spanish-distsim.tagger";
        } else if (lang.equals("chinese")) {
            posModel = "edu/stanford/nlp/models/pos-tagger/chinese-distsim/chinese-distsim.tagger";
        } else {
            posModel = "edu/stanford.nlp/models/pos-tagger/english-left3words/english-left3words-distsim.tagger";
        }
    } else {
        posModel = pm;
    }
    if (dm == null) {
        if (lang.equals("english")) {
            depModel = "edu/stanford/nlp/models/lexparser/englishRNN.ser.gz";
        } else if (lang.equals("spanish")) {
            depModel = "edu/stanford/nlp/models/lexparser/spanishPCFG.ser.gz";
        } else if (lang.equals("chinese")) {
            depModel = "edu/stanford/nlp/models/lexparser/chinesePCFG.ser.gz";
        } else {
            depModel = "edu/stanford/nlp/models/lexparser/englishRNN.ser.gz";
        }
    } else {
        depModel = dm;
    }

    //Create base properties:
    Properties props = new Properties();
    if (this.requiresDepCounts) {
        props.setProperty("annotators", "tokenize, ssplit, pos, parse");
    } else {
        props.setProperty("annotators", "tokenize, ssplit, pos");
    }

    //Create base pipeline:
    pipeline = new StanfordCoreNLP(props);

    try {
        //Create pipeline object:
        tokenizer = new TokenizerAnnotator(true, TokenizerAnnotator.TokenizerType.Whitespace);

        //Add objects to the pipeline:
        pipeline.addAnnotator(tokenizer);
    } catch (Exception ex) {
        System.out.println("ERROR: Problem while creating Stanford tokenizer.");
    }

    try {
        //Create pipeline object:
        tagger = new POSTaggerAnnotator(posModel, false);

        //Add object to the pipeline:
        pipeline.addAnnotator(tagger);
    } catch (Exception ex) {
        System.out.println(
                "ERROR: Problem while creating Stanford POS tagger. Please review the model paths and check for library availability.");
    }

    //If dependency counts are required:
    if (this.requiresDepCounts) {
        try {
            //Create pipeline object:
            parser = new ParserAnnotator(depModel, false, 300, StringUtils.EMPTY_STRING_ARRAY);

            //Add object to the pipeline:
            pipeline.addAnnotator(parser);
        } catch (Exception ex) {
            System.out.println(
                    "ERROR: Problem while creating Stanford dependency parser. Please review the model paths and check for library availability.");
        }
    }
}