Package org.carrot2.clustering.lingo

Class SimpleLabelAssigner

java.lang.Object

org.carrot2.attrs.AttrComposite

org.carrot2.clustering.lingo.SimpleLabelAssigner

All Implemented Interfaces:

AcceptingVisitor, LabelAssigner

public class SimpleLabelAssigner
extends AttrComposite
implements LabelAssigner

A simple and fast label assigner. For each base vector chooses the label that maximizes the base vector--label term vector cosine similarity. Different vectors can get the same label assigned, which means the number of final labels (after duplicate removal) may be smaller than the number of base vectors on input.

See Also:

UniqueLabelAssigner

Field Summary

Fields inherited from class org.carrot2.attrs.AttrComposite

attributes

Constructor Summary

Constructors

Constructor	Description
SimpleLabelAssigner()

Method Summary

Modifier and Type	Method	Description
void	assignLabels(LingoProcessingContext context, org.carrot2.math.mahout.matrix.DoubleMatrix2D stemCos, com.carrotsearch.hppc.IntIntHashMap filteredRowToStemIndex, org.carrot2.math.mahout.matrix.DoubleMatrix2D phraseCos)	Assigns labels to base vectors found by the matrix factorization.

Methods inherited from class org.carrot2.attrs.AttrComposite

accept

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.carrot2.attrs.AcceptingVisitor

accept

Constructor Details

SimpleLabelAssigner

public SimpleLabelAssigner()

Method Details

assignLabels

public void assignLabels(LingoProcessingContext context, org.carrot2.math.mahout.matrix.DoubleMatrix2D stemCos, com.carrotsearch.hppc.IntIntHashMap filteredRowToStemIndex, org.carrot2.math.mahout.matrix.DoubleMatrix2D phraseCos)

Description copied from interface: LabelAssigner

Assigns labels to base vectors found by the matrix factorization. The results must be stored in the LingoProcessingContext.clusterLabelFeatureIndex and LingoProcessingContext.clusterLabelScore arrays.

Specified by:

assignLabels in interface LabelAssigner

Parameters:

context - contains all information about the current clustering request

stemCos - base vector -- single stems cosine matrix

filteredRowToStemIndex - mapping between row indices of stemCos and indices of stems in PreprocessingContext.allStems

phraseCos - base vector -- phrase cosine matrix