cat.lump.ir.sim.ml.esa

Class EsaGenerator

java.lang.Object

cat.lump.ir.sim.ml.esa.EsaGenerator

public class EsaGenerator
extends java.lang.Object

A class that allows for passing from a text (collection) into its ESA vector representation.

Since:

Dec 17, 2013

Version:

0.1

Author:

albarron

Constructor Summary

Constructors

Constructor and Description
EsaGenerator(java.io.File indexPath, java.lang.String language) Invokes an instance of the EsaGenerator by loading the index and the analyzer for the required language TODO whether the indexpath should be established by default depending on the language

Method Summary

Methods

Modifier and Type	Method and Description
Vector	computeVector(java.lang.String text) Computes the ESA vector representation for the given text.
protected void	finalize() TODO important: check when to call this process instead of trusting the garbage collector Once the process is over, the index is closed
int	getIndexDimension()
void	loadIndex() Loads the Lucene index (previously created) with the reference corpus.
void	setAnalyzer() Set the Lucene analyzer to use according to the given language
void	setIndexPath(java.io.File indexPath) Set the path to Lucene's index
void	setNormalisation(java.lang.Boolean norm) Determines whether the ESA vectors are going to be normalised.

Methods inherited from class java.lang.Object

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

Constructor Detail

EsaGenerator

public EsaGenerator(java.io.File indexPath,
            java.lang.String language)

Invokes an instance of the EsaGenerator by loading the index and the analyzer for the required language

TODO whether the indexpath should be established by default depending on the language

Parameters:

indexPath -

language -

Method Detail

setAnalyzer

public void setAnalyzer()

Set the Lucene analyzer to use according to the given language

Parameters:

myAnalyzer -

setIndexPath

public void setIndexPath(java.io.File indexPath)

Set the path to Lucene's index

Parameters:

index_path -

loadIndex

public void loadIndex()

Loads the Lucene index (previously created) with the reference corpus.

Throws:

java.io.IOException

getIndexDimension

public int getIndexDimension()

Returns:

the dimension of the index (henceforth, the vector)

computeVector

public Vector computeVector(java.lang.String text)

Computes the ESA vector representation for the given text.

Parameters:

text -

Returns:

setNormalisation

public void setNormalisation(java.lang.Boolean norm)

Determines whether the ESA vectors are going to be normalised.
If true, the vectors are normalised by the maximum score in order to range the values in [0,1].

Parameters:

norm -

finalize

protected void finalize()

TODO important: check when to call this process instead of trusting the garbage collector Once the process is over, the index is closed

Overrides:

finalize in class java.lang.Object