com.wcohen.ss.expt

Class Blocker

java.lang.Object

com.wcohen.ss.expt.Blocker

Direct Known Subclasses:

AbbreviationsBlocker, NullBlocker, TokenBlocker

public abstract class Blocker
extends Object

Produces candidate pairs from a MatchData structure, and provides access to those candidate pairs.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	Blocker.Pair Holds a pair of instances, with mutable distance between them.

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	clusterMode

Constructor Summary

Constructors

Constructor and Description
Blocker()

Method Summary

Modifier and Type	Method and Description
abstract void	block(MatchData data) Load matchdata and prepare it for production of candidate pairs.
protected int	countCorrectPairs(MatchData data) Compute number of correct pairs betwn src1 and src2, where src2>src1
abstract Blocker.Pair	getPair(int i) Get the i-th candidate pair, as produced from most recently block()-ed data
abstract int	numCorrectPairs() Return total number of correct pairs in the dataset.
void	setClusterMode(boolean flag) In clusterMode, consider pairings between instances from the same source.
void	setClusterMode(Boolean flag)
abstract int	size() Return number of candidate pairs, as produced from most recently block()-ed data

Methods inherited from class java.lang.Object

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

Field Detail

clusterMode

protected boolean clusterMode

Constructor Detail

Blocker

public Blocker()

Method Detail

block

public abstract void block(MatchData data)

Load matchdata and prepare it for production of candidate pairs.

getPair

public abstract Blocker.Pair getPair(int i)

Get the i-th candidate pair, as produced from most recently block()-ed data

size

public abstract int size()

Return number of candidate pairs, as produced from most recently block()-ed data

setClusterMode

public final void setClusterMode(boolean flag)

In clusterMode, consider pairings between instances from the same source. If clusterMode is false, only consider pairing between instances from different sources.

setClusterMode

public final void setClusterMode(Boolean flag)

numCorrectPairs

public abstract int numCorrectPairs()

Return total number of correct pairs in the dataset.

countCorrectPairs

protected int countCorrectPairs(MatchData data)

Compute number of correct pairs betwn src1 and src2, where src2>src1