Package opennlp.tools.langdetect

Class LanguageDetectorCrossValidator

java.lang.Object

opennlp.tools.langdetect.LanguageDetectorCrossValidator

public class LanguageDetectorCrossValidator
extends Object

Cross validator for LanguageDetector.

Constructor Summary

Constructors

Constructor	Description
LanguageDetectorCrossValidator(TrainingParameters mlParams, LanguageDetectorFactory factory, LanguageDetectorEvaluationMonitor... listeners)	Initializes a LanguageDetectorCrossValidator with the given parameters.

Method Summary

Modifier and Type	Method	Description
void	evaluate(ObjectStream<LanguageSample> samples, int nFolds)	Starts the evaluation.
double	getDocumentAccuracy()
long	getDocumentCount()

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LanguageDetectorCrossValidator

public LanguageDetectorCrossValidator(TrainingParameters mlParams,
                                      LanguageDetectorFactory factory,
                                      LanguageDetectorEvaluationMonitor... listeners)

Initializes a LanguageDetectorCrossValidator with the given parameters.

Parameters:

mlParams - The TrainingParameters for the context of cross validation.

factory - The LanguageDetectorFactory for creating related objects.

listeners - the evaluation listeners.

Method Detail

evaluate

public void evaluate(ObjectStream<LanguageSample> samples,
                     int nFolds)
              throws IOException

Starts the evaluation.

Parameters:

samples - The ObjectStream of samples to train and test with.

nFolds - Number of folds. It must be greater than zero.

Throws:

IOException - Thrown if IO errors occurred.

getDocumentAccuracy

public double getDocumentAccuracy()

Returns:

Retrieves the word accuracy for all iterations.

getDocumentCount

public long getDocumentCount()

Returns:

Retrieves the number of words which where validated over all iterations. The result is the amount of folds multiplied by the total number of words.