cat.lump.ir.sim.ml.esa

Class Esa

java.lang.Object

cat.lump.ir.sim.ml.esa.Esa

All Implemented Interfaces:

Similarity

Direct Known Subclasses:

SimilarityESA

public abstract class Esa
extends java.lang.Object
implements Similarity

Field Summary

Fields

Modifier and Type	Field and Description
protected EsaVectors	esaVectorsA Instance of an ESA vector for the set of documents A.
protected EsaVectors	esaVectorsB Instance of an ESA vector for the set of documents B (see description for esaVectors_A
protected static LumpLogger	log

Constructor Summary

Constructors

Constructor and Description
Esa()

Method Summary

Methods

Modifier and Type	Method and Description
void	computePairwiseSimilarities() Compute only similarities for the matrix diagonal
void	computeSimilarities() Computes the ESA-based similarities between the previously loaded documents.
double	computeSimilarity(java.lang.String idA, java.lang.String idB) Computes the similarity between two specific documents.
void	displaySimilarities() Prints a matrix including all the similarities
protected boolean	documentsExist(java.lang.String id_A, java.lang.String id_B) Checks whether both documents exist already in the corresponding vector.
protected void	exitError(java.lang.String message)
java.util.Map<java.lang.String,java.lang.Double>	getPairwiseSimilarities()
java.util.Map<java.lang.String,java.util.Map<java.lang.String,java.lang.Double>>	getSimilarities()
Matrix	getSimilaritiesMatrix()
double	getSimilarity(java.lang.String idA)
double	getSimilarity(java.lang.String idA, java.lang.String idB) Obtains the similarity between texts id_A and id_B.

Methods inherited from class java.lang.Object

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

Field Detail

log

protected static LumpLogger log

esaVectorsA

protected EsaVectors esaVectorsA

Instance of an ESA vector for the set of documents A. It contains the representation of this documents on the basis of the reference corpus

esaVectorsB

protected EsaVectors esaVectorsB

Instance of an ESA vector for the set of documents B (see description for esaVectors_A

Constructor Detail

Esa

public Esa()

Method Detail

computeSimilarities

public void computeSimilarities()

Computes the ESA-based similarities between the previously loaded documents. The results are stored in a private variable which can be accessed either with getSimilarity() for a specific pair or with getSimilarities() for a nested map including all the similarities for every document pair.

Specified by:

computeSimilarities in interface Similarity

computeSimilarity

public double computeSimilarity(java.lang.String idA,
                       java.lang.String idB)

Computes the similarity between two specific documents. (if previously computed, getSimilarity() can be simply called.

Specified by:

computeSimilarity in interface Similarity

Parameters:

idA -

idB -

Returns:

sim(idA, idB) if both documents exist; -1 otherwise

getSimilarity

public double getSimilarity(java.lang.String idA,
                   java.lang.String idB)

Obtains the similarity between texts id_A and id_B. The documents should have been loaded before and similarities computed, both through computeSimilarities()

Specified by:

getSimilarity in interface Similarity

Parameters:

idA -

idB -

Returns:

sim(id_A, id_B); -1 if not computed beforehand.

displaySimilarities

public void displaySimilarities()

Description copied from interface: Similarity

Prints a matrix including all the similarities

Specified by:

displaySimilarities in interface Similarity

documentsExist

protected boolean documentsExist(java.lang.String id_A,
                     java.lang.String id_B)

Checks whether both documents exist already in the corresponding vector.

Parameters:

id_A -

id_B -

Returns:

true if both documents were previously loaded

computePairwiseSimilarities

public void computePairwiseSimilarities()

Compute only similarities for the matrix diagonal

getSimilaritiesMatrix

public Matrix getSimilaritiesMatrix()

Returns:

a matrix with all the computed similarities

getSimilarities

public java.util.Map<java.lang.String,java.util.Map<java.lang.String,java.lang.Double>> getSimilarities()

Specified by:

getSimilarities in interface Similarity

Returns:

a nested map with all the computed similarities.

getPairwiseSimilarities

public java.util.Map<java.lang.String,java.lang.Double> getPairwiseSimilarities()

Returns:

nested map containing the pairwise similarities.

getSimilarity

public double getSimilarity(java.lang.String idA)

Parameters:

idA -

Returns:

exitError

protected void exitError(java.lang.String message)