org.apache.lucene.search.similarities

Class DFRSimilarity

java.lang.Object

org.apache.lucene.search.similarities.Similarity

org.apache.lucene.search.similarities.SimilarityBase

org.apache.lucene.search.similarities.DFRSimilarity

public class DFRSimilarity
extends SimilarityBase

Implements the divergence from randomness (DFR) framework introduced in Gianni Amati and Cornelis Joost Van Rijsbergen. 2002. Probabilistic models of information retrieval based on measuring the divergence from randomness. ACM Trans. Inf. Syst. 20, 4 (October 2002), 357-389.

The DFR scoring formula is composed of three separate components: the basic model, the aftereffect and an additional normalization component, represented by the classes BasicModel, AfterEffect and Normalization, respectively. The names of these classes were chosen to match the names of their counterparts in the Terrier IR engine.

To construct a DFRSimilarity, you must specify the implementations for all three components of DFR:

1. BasicModel: Basic model of information content:
   • BasicModelG: Geometric approximation of Bose-Einstein
   • BasicModelIn: Inverse document frequency
   • BasicModelIne: Inverse expected document frequency [mixture of Poisson and IDF]
   • BasicModelIF: Inverse term frequency [approximation of I(ne)]
2. AfterEffect: First normalization of information gain:
   • AfterEffectL: Laplace's law of succession
   • AfterEffectB: Ratio of two Bernoulli processes
3. Normalization: Second (length) normalization:
   • NormalizationH1: Uniform distribution of term frequency
   • NormalizationH2: term frequency density inversely related to length
   • NormalizationH3: term frequency normalization provided by Dirichlet prior
   • NormalizationZ: term frequency normalization provided by a Zipfian relation
   • Normalization.NoNormalization: no second normalization

Note that qtf, the multiplicity of term-occurrence in the query, is not handled by this implementation.

Note that basic models BE (Limiting form of Bose-Einstein), P (Poisson approximation of the Binomial) and D (Divergence approximation of the Binomial) are not implemented because their formula couldn't be written in a way that makes scores non-decreasing with the normalized term frequency.

See Also:

BasicModel, AfterEffect, Normalization

WARNING: This API is experimental and might change in incompatible ways in the next release.

Nested Class Summary

Nested classes/interfaces inherited from class org.apache.lucene.search.similarities.Similarity

Similarity.SimScorer

Field Summary

Fields

Modifier and Type	Field and Description
protected AfterEffect	afterEffect The first normalization of the information content.
protected BasicModel	basicModel The basic model for information content.
protected Normalization	normalization The term frequency normalization.

Fields inherited from class org.apache.lucene.search.similarities.SimilarityBase

discountOverlaps

Constructor Summary

Constructors

Constructor and Description
DFRSimilarity(BasicModel basicModel, AfterEffect afterEffect, Normalization normalization) Creates DFRSimilarity from the three components.

Method Summary

Modifier and Type	Method and Description
protected Explanation	explain(BasicStats stats, Explanation freq, double docLen) Explains the score.
protected void	explain(List<Explanation> subs, BasicStats stats, double freq, double docLen) Subclasses should implement this method to explain the score.
AfterEffect	getAfterEffect() Returns the first normalization
BasicModel	getBasicModel() Returns the basic model of information content
Normalization	getNormalization() Returns the second normalization
protected double	score(BasicStats stats, double freq, double docLen) Scores the document doc.
String	toString() Subclasses must override this method to return the name of the Similarity and preferably the values of parameters (if any) as well.

Methods inherited from class org.apache.lucene.search.similarities.SimilarityBase

computeNorm, fillBasicStats, getDiscountOverlaps, log2, newStats, scorer, setDiscountOverlaps

Methods inherited from class java.lang.Object

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

Field Detail

basicModel

protected final BasicModel basicModel

The basic model for information content.

afterEffect

protected final AfterEffect afterEffect

The first normalization of the information content.

normalization

protected final Normalization normalization

The term frequency normalization.

Constructor Detail

DFRSimilarity

public DFRSimilarity(BasicModel basicModel,
                     AfterEffect afterEffect,
                     Normalization normalization)

Creates DFRSimilarity from the three components.

Note that null values are not allowed: if you want no normalization, instead pass Normalization.NoNormalization.

Parameters:

basicModel - Basic model of information content

afterEffect - First normalization of information gain

normalization - Second (length) normalization

Method Detail

score

protected double score(BasicStats stats,
                       double freq,
                       double docLen)

Description copied from class: SimilarityBase

Scores the document doc.

Subclasses must apply their scoring formula in this class.

Specified by:

score in class SimilarityBase

Parameters:

stats - the corpus level statistics.

freq - the term frequency.

docLen - the document length.

Returns:

the score.

explain

protected void explain(List<Explanation> subs,
                       BasicStats stats,
                       double freq,
                       double docLen)

Description copied from class: SimilarityBase

Subclasses should implement this method to explain the score. expl already contains the score, the name of the class and the doc id, as well as the term frequency and its explanation; subclasses can add additional clauses to explain details of their scoring formulae.

The default implementation does nothing.

Overrides:

explain in class SimilarityBase

Parameters:

subs - the list of details of the explanation to extend

stats - the corpus level statistics.

freq - the term frequency.

docLen - the document length.

explain

protected Explanation explain(BasicStats stats,
                              Explanation freq,
                              double docLen)

Description copied from class: SimilarityBase

Explains the score. The implementation here provides a basic explanation in the format score(name-of-similarity, doc=doc-id, freq=term-frequency), computed from:, and attaches the score (computed via the SimilarityBase.score(BasicStats, double, double) method) and the explanation for the term frequency. Subclasses content with this format may add additional details in SimilarityBase.explain(List, BasicStats, double, double).

Overrides:

explain in class SimilarityBase

Parameters:

stats - the corpus level statistics.

freq - the term frequency and its explanation.

docLen - the document length.

Returns:

the explanation.

toString

public String toString()

Description copied from class: SimilarityBase

Subclasses must override this method to return the name of the Similarity and preferably the values of parameters (if any) as well.

Specified by:

toString in class SimilarityBase

getBasicModel

public BasicModel getBasicModel()

Returns the basic model of information content

getAfterEffect

public AfterEffect getAfterEffect()

Returns the first normalization

getNormalization

public Normalization getNormalization()

Returns the second normalization