Package org.carrot2.clustering.kmeans

Class BisectingKMeansClusteringAlgorithm

java.lang.Object

org.carrot2.attrs.AttrComposite

org.carrot2.clustering.kmeans.BisectingKMeansClusteringAlgorithm

All Implemented Interfaces:

AcceptingVisitor, ClusteringAlgorithm

public class BisectingKMeansClusteringAlgorithm
extends AttrComposite
implements ClusteringAlgorithm

A very simple implementation of bisecting k-means clustering. Unlike other algorithms in Carrot2, this one creates hard clustering (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	Description
AttrInteger	clusterCount	The number of clusters to create.
AttrInteger	labelCount	Label count.
TermDocumentMatrixBuilder	matrixBuilder	Term-document matrix builder for the algorithm.
TermDocumentMatrixReducer	matrixReducer	Term-document matrix reducer for the algorithm.
AttrInteger	maxIterations	The maximum number of k-means iterations to perform.
static String	NAME
AttrInteger	partitionCount	Partition count.
BasicPreprocessingPipeline	preprocessing	A pipeline of components transforming input documents into a PreprocessingContext.
AttrString	queryHint	Query terms used to retrieve documents.
AttrBoolean	useDimensionalityReduction	Use dimensionality reduction.

Fields inherited from class org.carrot2.attrs.AttrComposite

attributes

Constructor Summary

Constructors

Constructor	Description
BisectingKMeansClusteringAlgorithm()

Method Summary

Modifier and Type	Method	Description
<T extends Document> List<Cluster<T>>	cluster(Stream<? extends T> docStream, LanguageComponents languageComponents)
Set<Class<?>>	requiredLanguageComponents()

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

Methods inherited from interface org.carrot2.clustering.ClusteringAlgorithm

supports

Field Details

NAME

public static final String NAME

See Also:

Constant Field Values

clusterCount

public final AttrInteger clusterCount

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

maxIterations

public final AttrInteger maxIterations

The maximum number of k-means iterations to perform.

partitionCount

public final AttrInteger partitionCount

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

labelCount

public final AttrInteger labelCount

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

queryHint

public final AttrString queryHint

Query terms used to retrieve documents. The query is used as a hint to avoid trivial clusters.

useDimensionalityReduction

public final AttrBoolean 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.

matrixBuilder

public TermDocumentMatrixBuilder matrixBuilder

Term-document matrix builder for the algorithm.

matrixReducer

public TermDocumentMatrixReducer matrixReducer

Term-document matrix reducer for the algorithm.

preprocessing

public BasicPreprocessingPipeline preprocessing

A pipeline of components transforming input documents into a PreprocessingContext.

Constructor Details

BisectingKMeansClusteringAlgorithm

public BisectingKMeansClusteringAlgorithm()

Method Details

requiredLanguageComponents

public Set<Class<?>> requiredLanguageComponents()

Specified by:

requiredLanguageComponents in interface ClusteringAlgorithm

cluster

public <T extends Document> List<Cluster<T>> cluster(Stream<? extends T> docStream, LanguageComponents languageComponents)

Specified by:

cluster in interface ClusteringAlgorithm