com.github.pmerienne.trident.ml.nlp

Class KLDClassifier

java.lang.Object

com.github.pmerienne.trident.ml.nlp.KLDClassifier

All Implemented Interfaces:

TextClassifier<Integer>, TextFeaturesExtractor, Serializable

public class KLDClassifier
extends Object
implements TextClassifier<Integer>, TextFeaturesExtractor, Serializable

Text classifier and feature extractor using Kullback-Leibler Distance. See "Using Kullback-Leibler Distance for Text Categorization Brigitte Bigi"

Author:

pmerienne

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
KLDClassifier()
KLDClassifier(int nbClasses)
KLDClassifier(int nbClasses, int maxWordsPerClass)
KLDClassifier(int nbClasses, int maxWordsPerClass, double thresholdFactor, boolean normalize)

Method Summary

Methods

Modifier and Type	Method and Description
protected double	caculateBeta(Vocabulary documentVocabulary)
protected double	calculateGamma(int classIndex)
Integer	classify(List<String> documentWords)
double[]	distance(List<String> documentWords)
protected Double	distance(Vocabulary documentVocabulary, int classIndex, double beta)
protected double	estimateEpsilon()
double[]	extractFeatures(List<String> documentWords)
protected double	getEpsilon()
protected double	getGamma(int classIndex)
void	update(Integer classIndex, List<String> documentWords)
protected Double	wordProbabilityInCategory(String word, int classIndex)
protected Double	wordProbabilityInDocument(String word, Vocabulary documentVocabulary, double beta)

Methods inherited from class java.lang.Object

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

Constructor Detail

KLDClassifier

public KLDClassifier()

KLDClassifier

public KLDClassifier(int nbClasses)

KLDClassifier

public KLDClassifier(int nbClasses,
             int maxWordsPerClass)

KLDClassifier

public KLDClassifier(int nbClasses,
             int maxWordsPerClass,
             double thresholdFactor,
             boolean normalize)

Method Detail

extractFeatures

public double[] extractFeatures(List<String> documentWords)

Specified by:

extractFeatures in interface TextFeaturesExtractor

update

public void update(Integer classIndex,
          List<String> documentWords)

Specified by:

update in interface TextClassifier<Integer>

classify

public Integer classify(List<String> documentWords)

Specified by:

classify in interface TextClassifier<Integer>

distance

public double[] distance(List<String> documentWords)

distance

protected Double distance(Vocabulary documentVocabulary,
              int classIndex,
              double beta)

wordProbabilityInCategory

protected Double wordProbabilityInCategory(String word,
                               int classIndex)

wordProbabilityInDocument

protected Double wordProbabilityInDocument(String word,
                               Vocabulary documentVocabulary,
                               double beta)

getEpsilon

protected double getEpsilon()

estimateEpsilon

protected double estimateEpsilon()

getGamma

protected double getGamma(int classIndex)

calculateGamma

protected double calculateGamma(int classIndex)

caculateBeta

protected double caculateBeta(Vocabulary documentVocabulary)