Search in sources :

Example 1 with TemporalChunkerAnnotator

use of edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator in project cogcomp-nlp by CogComp.

the class TestTemporalChunker method setUp.

@Before
public void setUp() throws IOException, ParserConfigurationException, SAXException {
    Properties rmProps = new TemporalChunkerConfigurator().getDefaultConfig().getProperties();
    rmProps.setProperty("useHeidelTime", "False");
    tca = new TemporalChunkerAnnotator(new ResourceManager(rmProps));
}
Also used : TemporalChunkerConfigurator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerConfigurator) TemporalChunkerAnnotator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator) ResourceManager(edu.illinois.cs.cogcomp.core.utilities.configuration.ResourceManager) Properties(java.util.Properties) Before(org.junit.Before)

Example 2 with TemporalChunkerAnnotator

use of edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator in project cogcomp-nlp by CogComp.

the class PipelineFactory method buildAnnotators.

/**
 * 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.
 *
 * @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);
        // per sentence? could be per
        stanfordProps.put("parse.maxtime", timePerSentence);
        // 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;
}
Also used : HashMap(java.util.HashMap) VerbSenseAnnotator(edu.illinois.cs.cogcomp.verbsense.VerbSenseAnnotator) MentionAnnotator(org.cogcomp.md.MentionAnnotator) Properties(java.util.Properties) POSTaggerAnnotator(edu.stanford.nlp.pipeline.POSTaggerAnnotator) PrepSRLAnnotator(edu.illinois.cs.cogcomp.prepsrl.PrepSRLAnnotator) Language(edu.illinois.cs.cogcomp.core.constants.Language) SemanticRoleLabeler(edu.illinois.cs.cogcomp.srl.SemanticRoleLabeler) TemporalChunkerAnnotator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator) W2VDatalessAnnotator(edu.illinois.cs.cogcomp.datalessclassification.ta.W2VDatalessAnnotator) TemporalChunkerConfigurator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerConfigurator) Stanford331Configurator(edu.illinois.cs.cogcomp.pipeline.common.Stanford331Configurator) ParserAnnotator(edu.stanford.nlp.pipeline.ParserAnnotator) ESADatalessAnnotator(edu.illinois.cs.cogcomp.datalessclassification.ta.ESADatalessAnnotator) POSAnnotator(edu.illinois.cs.cogcomp.pos.POSAnnotator) ResourceManager(edu.illinois.cs.cogcomp.core.utilities.configuration.ResourceManager) IOException(java.io.IOException) RelationAnnotator(org.cogcomp.re.RelationAnnotator) IOException(java.io.IOException) TemporalChunkerAnnotator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator) ChunkerAnnotator(edu.illinois.cs.cogcomp.chunker.main.ChunkerAnnotator) PipelineConfigurator(edu.illinois.cs.cogcomp.pipeline.common.PipelineConfigurator) TemporalChunkerAnnotator(edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator) W2VDatalessAnnotator(edu.illinois.cs.cogcomp.datalessclassification.ta.W2VDatalessAnnotator) RelationAnnotator(org.cogcomp.re.RelationAnnotator) ESADatalessAnnotator(edu.illinois.cs.cogcomp.datalessclassification.ta.ESADatalessAnnotator) POSTaggerAnnotator(edu.stanford.nlp.pipeline.POSTaggerAnnotator) PrepSRLAnnotator(edu.illinois.cs.cogcomp.prepsrl.PrepSRLAnnotator) ParserAnnotator(edu.stanford.nlp.pipeline.ParserAnnotator) MentionAnnotator(org.cogcomp.md.MentionAnnotator) QuestionTypeAnnotator(edu.illinois.cs.cogcomp.question_typer.QuestionTypeAnnotator) ChunkerAnnotator(edu.illinois.cs.cogcomp.chunker.main.ChunkerAnnotator) DepAnnotator(edu.illinois.cs.cogcomp.depparse.DepAnnotator) POSAnnotator(edu.illinois.cs.cogcomp.pos.POSAnnotator) VerbSenseAnnotator(edu.illinois.cs.cogcomp.verbsense.VerbSenseAnnotator) NERAnnotator(edu.illinois.cs.cogcomp.ner.NERAnnotator) NERAnnotator(edu.illinois.cs.cogcomp.ner.NERAnnotator) ESADatalessConfigurator(edu.illinois.cs.cogcomp.datalessclassification.config.ESADatalessConfigurator) DepAnnotator(edu.illinois.cs.cogcomp.depparse.DepAnnotator) W2VDatalessConfigurator(edu.illinois.cs.cogcomp.datalessclassification.config.W2VDatalessConfigurator) IllinoisLemmatizer(edu.illinois.cs.cogcomp.nlp.lemmatizer.IllinoisLemmatizer) Quantifier(edu.illinois.cs.cogcomp.quant.driver.Quantifier) CommaLabeler(edu.illinois.cs.cogcomp.comma.CommaLabeler) StanfordParseHandler(edu.illinois.cs.cogcomp.pipeline.handlers.StanfordParseHandler) QuestionTypeAnnotator(edu.illinois.cs.cogcomp.question_typer.QuestionTypeAnnotator) StanfordDepHandler(edu.illinois.cs.cogcomp.pipeline.handlers.StanfordDepHandler)

Aggregations

ResourceManager (edu.illinois.cs.cogcomp.core.utilities.configuration.ResourceManager)2 TemporalChunkerAnnotator (edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerAnnotator)2 TemporalChunkerConfigurator (edu.illinois.cs.cogcomp.temporal.normalizer.main.TemporalChunkerConfigurator)2 Properties (java.util.Properties)2 ChunkerAnnotator (edu.illinois.cs.cogcomp.chunker.main.ChunkerAnnotator)1 CommaLabeler (edu.illinois.cs.cogcomp.comma.CommaLabeler)1 Language (edu.illinois.cs.cogcomp.core.constants.Language)1 ESADatalessConfigurator (edu.illinois.cs.cogcomp.datalessclassification.config.ESADatalessConfigurator)1 W2VDatalessConfigurator (edu.illinois.cs.cogcomp.datalessclassification.config.W2VDatalessConfigurator)1 ESADatalessAnnotator (edu.illinois.cs.cogcomp.datalessclassification.ta.ESADatalessAnnotator)1 W2VDatalessAnnotator (edu.illinois.cs.cogcomp.datalessclassification.ta.W2VDatalessAnnotator)1 DepAnnotator (edu.illinois.cs.cogcomp.depparse.DepAnnotator)1 NERAnnotator (edu.illinois.cs.cogcomp.ner.NERAnnotator)1 IllinoisLemmatizer (edu.illinois.cs.cogcomp.nlp.lemmatizer.IllinoisLemmatizer)1 PipelineConfigurator (edu.illinois.cs.cogcomp.pipeline.common.PipelineConfigurator)1 Stanford331Configurator (edu.illinois.cs.cogcomp.pipeline.common.Stanford331Configurator)1 StanfordDepHandler (edu.illinois.cs.cogcomp.pipeline.handlers.StanfordDepHandler)1 StanfordParseHandler (edu.illinois.cs.cogcomp.pipeline.handlers.StanfordParseHandler)1 POSAnnotator (edu.illinois.cs.cogcomp.pos.POSAnnotator)1 PrepSRLAnnotator (edu.illinois.cs.cogcomp.prepsrl.PrepSRLAnnotator)1