org.carrot2.clustering.kmeans

Class BisectingKMeansClusteringAlgorithm

java.lang.Object

org.carrot2.core.ProcessingComponentBase

org.carrot2.clustering.kmeans.BisectingKMeansClusteringAlgorithm

All Implemented Interfaces:

IClusteringAlgorithm, IProcessingComponent

@Bindable(prefix="BisectingKMeansClusteringAlgorithm",
          inherit=CommonAttributes.class)
public class BisectingKMeansClusteringAlgorithm
extends ProcessingComponentBase
implements IClusteringAlgorithm

A very simple implementation of bisecting k-means clustering. Unlike other algorithms in Carrot2, this one creates hard clusterings (one document belongs only to one cluster). On the other hand, the clusters are labeled only with individual words that may not always fully correspond to all documents in the cluster.

Field Summary

Fields

Modifier and Type	Field and Description
int	clusterCount The number of clusters to create.
List<Cluster>	clusters
List<Document>	documents
int	labelCount Label count.
LabelFormatter	labelFormatter Cluster label formatter, contains bindable attributes.
TermDocumentMatrixBuilder	matrixBuilder Term-document matrix builder for the algorithm, contains bindable attributes.
TermDocumentMatrixReducer	matrixReducer Term-document matrix reducer for the algorithm, contains bindable attributes.
int	maxIterations The maximum number of k-means iterations to perform.
MultilingualClustering	multilingualClustering A helper for performing multilingual clustering.
int	partitionCount Partition count.
IPreprocessingPipeline	preprocessingPipeline Common preprocessing tasks handler.
boolean	useDimensionalityReduction Use dimensionality reduction.

Constructor Summary

Constructors

Constructor and Description
BisectingKMeansClusteringAlgorithm()

Method Summary

Modifier and Type	Method and Description
protected void	cluster(LanguageCode language) Perform clustering for a given language.
void	process() Performs the processing required to fulfill the request.

Methods inherited from class org.carrot2.core.ProcessingComponentBase

afterProcessing, beforeProcessing, dispose, getContext, getSharedExecutor, init

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.carrot2.core.IProcessingComponent

afterProcessing, beforeProcessing, dispose, init

Field Detail

documents

@Processing
 @Input
 @Required
 @Internal
 @Attribute(key="documents",
           inherit=true)
public List<Document> documents

clusters

@Processing
 @Output
 @Internal
 @Attribute(key="clusters",
           inherit=true)
public List<Cluster> clusters

clusterCount

@Processing
 @Input
 @Attribute
 @IntRange(min=2)
 @Group(value="Clusters")
 @Level(value=BASIC)
 @Label(value="Cluster count")
public int clusterCount

The number of clusters to create. The algorithm will create at most the specified number of clusters.

maxIterations

@Processing
 @Input
 @Attribute
 @IntRange(min=1)
 @Group(value="K-means")
 @Level(value=BASIC)
 @Label(value="Maximum iterations")
public int maxIterations

The maximum number of k-means iterations to perform.

useDimensionalityReduction

@Processing
 @Input
 @Attribute
 @Group(value="K-means")
 @Level(value=BASIC)
 @Label(value="Use dimensionality reduction")
public boolean useDimensionalityReduction

Use dimensionality reduction. If true, k-means will be applied on the dimensionality-reduced term-document matrix with the number of dimensions being equal to twice the number of requested clusters. If the number of dimensions is lower than the number of input documents, reduction will not be performed. If false, the k-means will be performed directly on the original term-document matrix.

partitionCount

@Processing
 @Input
 @Attribute
 @IntRange(min=2,
          max=10)
 @Group(value="K-means")
 @Level(value=BASIC)
 @Label(value="Partition count")
public int partitionCount

Partition count. The number of partitions to create at each k-means clustering iteration.

labelCount

@Processing
 @Input
 @Attribute
 @IntRange(min=1,
          max=10)
 @Group(value="Clusters")
 @Level(value=BASIC)
 @Label(value="Label count")
public int labelCount

Label count. The minimum number of labels to return for each cluster.

preprocessingPipeline

@Input
 @Attribute
 @Internal
 @Level(value=ADVANCED)
public IPreprocessingPipeline preprocessingPipeline

Common preprocessing tasks handler.

matrixBuilder

public final TermDocumentMatrixBuilder matrixBuilder

Term-document matrix builder for the algorithm, contains bindable attributes.

matrixReducer

public final TermDocumentMatrixReducer matrixReducer

Term-document matrix reducer for the algorithm, contains bindable attributes.

labelFormatter

public final LabelFormatter labelFormatter

Cluster label formatter, contains bindable attributes.

multilingualClustering

public final MultilingualClustering multilingualClustering

A helper for performing multilingual clustering.

Constructor Detail

BisectingKMeansClusteringAlgorithm

public BisectingKMeansClusteringAlgorithm()

Method Detail

process

public void process()
             throws ProcessingException

Description copied from interface: IProcessingComponent

Performs the processing required to fulfill the request. This method is called once per request.

Specified by:

process in interface IProcessingComponent

Overrides:

process in class ProcessingComponentBase

Throws:

ProcessingException - when processing failed. If thrown, the IProcessingComponent.afterProcessing() method will be called and the component will be ready to accept further requests or to be disposed of. Finally, the exception will be rethrown from the controller method that caused the component to perform processing.

cluster

protected void cluster(LanguageCode language)

Perform clustering for a given language.