Example usage for edu.stanford.nlp.sequences ColumnDocumentReaderAndWriter ColumnDocumentReaderAndWriter

Introduction

In this page you can find the example usage for edu.stanford.nlp.sequences ColumnDocumentReaderAndWriter ColumnDocumentReaderAndWriter.

Prototype

ColumnDocumentReaderAndWriter

Source Link

Usage

From source file:org.exist.xquery.corenlp.TrainClassifier.java

License:Open Source License

private void trainClassifier(Collection<List<CoreLabel>> documents, final InputDocType inputFormat) {
    final Properties props = new Properties();
    // fixme! - check ocrTrain configurable under other name?
    //props.setProperty("ocrTrain", "true");
    //props.setProperty("serializeTo", tempOutFile.toAbsolutePath().toString());
    props.setProperty("useClassFeature", "true");
    props.setProperty("useWord", "true");
    props.setProperty("useNGrams", "true");
    props.setProperty("noMidNGrams", "true");
    props.setProperty("useDisjunctive", "true");
    props.setProperty("maxNGramLeng", "6");
    props.setProperty("usePrev", "true");
    props.setProperty("useNext", "true");
    props.setProperty("useSequences", "true");
    props.setProperty("usePrevSequences", "true");
    props.setProperty("maxLeft", "1");
    props.setProperty("useTypeSeqs", "true");
    props.setProperty("useTypeSeqs2", "true");
    props.setProperty("useTypeySequences", "true");
    props.setProperty("wordShape", "chris2useLC");

    CRFClassifier<CoreLabel> classifier = new CRFClassifier(props);
    classifier.train(documents, new ColumnDocumentReaderAndWriter());
    classifier.serializeClassifier(tempOutFile.toAbsolutePath().toString());
}

---