cat.lump.ir.lucene.query

Class Document2Query

java.lang.Object

cat.lump.ir.lucene.query.Document2Query

public class Document2Query
extends java.lang.Object

The contents of a document are processed to be in the right format for Lucene querying.
In order to do that, a tokenizer is called that acts exactly as it did during the index generation depends on Lucene's Analyzers) TODO determine which format is better for different interests, such as ESA TODO the stemming process seems to be missing here. Check if when querying it is carriend out

Since:

April 12 2012

Author:

albarron

Constructor Summary

Constructors

Constructor and Description
Document2Query()
Document2Query(java.util.Locale lan)

Method Summary

Methods

Modifier and Type	Method and Description
java.lang.String	doc2WeightQuery(java.lang.String file) Generates a query in which tokens' relevance depend on their frequency
java.lang.String	file2FlatQuery(java.lang.String file) Generates a query in which every token has the same relevance TODO why am I using the same tokenizer for every language???
static java.lang.String	flatQuery(java.lang.String[] tokens) Creates a query considering all the tokens (i.e. some words could be repeated)
Analyzer	getAnalyzer()
java.lang.String	str2FlatQuery(Analyzer analyzer, java.lang.String text) Generates a query in which every token has the same relevance
java.lang.String	str2FlatQuery(java.lang.String text) Generates a query in which every token has the same relevance
static java.lang.String	vocQuery(java.lang.String[] tokens) Creates a query considering only the vocabulary (i.e. types)
static java.lang.String	weightQuery(java.lang.String[] tokens) Creates a query where the relevance of a type depends on its frequency (i.e. if a token w appears 4 times, it will appear as w^4)

Methods inherited from class java.lang.Object

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

Constructor Detail

Document2Query

public Document2Query()

Document2Query

public Document2Query(java.util.Locale lan)

Method Detail

getAnalyzer

public Analyzer getAnalyzer()

file2FlatQuery

public java.lang.String file2FlatQuery(java.lang.String file)

Generates a query in which every token has the same relevance TODO why am I using the same tokenizer for every language???

Parameters:

file -

Returns:

string representation of the query

Throws:

java.io.IOException

str2FlatQuery

public java.lang.String str2FlatQuery(Analyzer analyzer,
                             java.lang.String text)

Generates a query in which every token has the same relevance

Parameters:

analyzer -

text - string representation of the query

Returns:

string with space-separated tokens

str2FlatQuery

public java.lang.String str2FlatQuery(java.lang.String text)

Generates a query in which every token has the same relevance

Parameters:

analyzer -

text - string representation of the query

Returns:

string with space-separated tokens

doc2WeightQuery

public java.lang.String doc2WeightQuery(java.lang.String file)

Generates a query in which tokens' relevance depend on their frequency

Parameters:

file -

Returns:

a string with space-separated tokens and weights

vocQuery

public static java.lang.String vocQuery(java.lang.String[] tokens)

Creates a query considering only the vocabulary (i.e. types)

Parameters:

tokens -

Returns:

a string with space-separated tokens

flatQuery

public static java.lang.String flatQuery(java.lang.String[] tokens)

Creates a query considering all the tokens (i.e. some words could be repeated)

Parameters:

tokens -

Returns:

string with flat query (space-separated tokens)

weightQuery

public static java.lang.String weightQuery(java.lang.String[] tokens)

Creates a query where the relevance of a type depends on its frequency (i.e. if a token w appears 4 times, it will appear as w^4)

Parameters:

tokens -

Returns:

string where tokens are weighted by their frequency