eu.project.ttc.models

Class CrossTable

java.lang.Object

eu.project.ttc.models.CrossTable

public class CrossTable
extends java.lang.Object

An index (often called cross table) that helps to compute the association rate of two terms, i.e. their amount of co-occurrency in the corpus.

Constructor Summary

Constructors

Constructor and Description
CrossTable(TermIndex termIndex) Constructs this cross table with a term index and inits the table with all co-occurrences of the term index's terms.

Method Summary

Modifier and Type	Method and Description
double	computeRate(AssociationRate rate, Term x, Term y) Computes coefficients a, b, c and d (available) and computes the association rate based on these coefficients.
int	getLastA() The number of times coTerm id a co-occurrence of term WARNING : not thread safe !
int	getLastB() The number of times a co-occurrence of term is not coTerm WARNING : not thread safe !
int	getLastC() The number of times coTerm is a co-occurrence of another term than term.
int	getLastD() The number of times neither term nor coTerm is a co-occurrence of something else WARNING : not thread safe !

Methods inherited from class java.lang.Object

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

Constructor Detail

CrossTable

public CrossTable(TermIndex termIndex)

Constructs this cross table with a term index and inits the table with all co-occurrences of the term index's terms. This init process is based on terms' context vectors, which must be computed before invoking this constructor.

Parameters:

termIndex - The term index of the corpus

Throws:

java.lang.NullPointerException - if terms' context vectors are not already computed

java.lang.NullPointerException - when the context vectors of term index terms are not

See Also:

Term.getContextVector()

Method Detail

computeRate

public double computeRate(AssociationRate rate,
                          Term x,
                          Term y)

Computes coefficients a, b, c and d (available) and computes the association rate based on these coefficients. These coefficients are made available through the not-thread safe methods #getLastX().

Parameters:

rate - the association rate measure

x - the base term

y - the co term

Returns:

the association rate

See Also:

getLastA(), getLastB(), getLastC(), getLastD()

getLastA

public int getLastA()

The number of times coTerm id a co-occurrence of term WARNING : not thread safe !

Returns:

getLastB

public int getLastB()

The number of times a co-occurrence of term is not coTerm WARNING : not thread safe !

Returns:

getLastC

public int getLastC()

The number of times coTerm is a co-occurrence of another term than term. WARNING : not thread safe !

Returns:

getLastD

public int getLastD()

The number of times neither term nor coTerm is a co-occurrence of something else WARNING : not thread safe !

Returns: