org.apache.lucene.search.similarities.DFRSimilarity Maven / Gradle / Ivy
Show all versions of aem-sdk-api Show documentation
/*
* COPIED FROM APACHE LUCENE 4.7.2
*
* Git URL: [email protected]:apache/lucene.git, tag: releases/lucene-solr/4.7.2, path: lucene/core/src/java
*
* (see https://issues.apache.org/jira/browse/OAK-10786 for details)
*/
package org.apache.lucene.search.similarities;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import org.apache.lucene.search.Explanation;
import org.apache.lucene.search.similarities.AfterEffect.NoAfterEffect;
import org.apache.lucene.search.similarities.Normalization.NoNormalization;
/**
* 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
* {@code BasicModel}, {@code AfterEffect} and {@code 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:
*
* - {@link BasicModel}: Basic model of information content:
*
* - {@link BasicModelBE}: Limiting form of Bose-Einstein
*
- {@link BasicModelG}: Geometric approximation of Bose-Einstein
*
- {@link BasicModelP}: Poisson approximation of the Binomial
*
- {@link BasicModelD}: Divergence approximation of the Binomial
*
- {@link BasicModelIn}: Inverse document frequency
*
- {@link BasicModelIne}: Inverse expected document
* frequency [mixture of Poisson and IDF]
*
- {@link BasicModelIF}: Inverse term frequency
* [approximation of I(ne)]
*
* - {@link AfterEffect}: First normalization of information
* gain:
*
* - {@link AfterEffectL}: Laplace's law of succession
*
- {@link AfterEffectB}: Ratio of two Bernoulli processes
*
- {@link NoAfterEffect}: no first normalization
*
* - {@link Normalization}: Second (length) normalization:
*
* - {@link NormalizationH1}: Uniform distribution of term
* frequency
*
- {@link NormalizationH2}: term frequency density inversely
* related to length
*
- {@link NormalizationH3}: term frequency normalization
* provided by Dirichlet prior
*
- {@link NormalizationZ}: term frequency normalization provided
* by a Zipfian relation
*
- {@link NoNormalization}: no second normalization
*
*
* Note that qtf, the multiplicity of term-occurrence in the query,
* is not handled by this implementation.
* @see BasicModel
* @see AfterEffect
* @see Normalization
* @lucene.experimental
*/
public class DFRSimilarity extends SimilarityBase {
/** The basic model for information content. */
protected final BasicModel basicModel;
/** The first normalization of the information content. */
protected final AfterEffect afterEffect;
/** The term frequency normalization. */
protected final Normalization normalization;
/**
* Creates DFRSimilarity from the three components.
*
* Note that null values are not allowed:
* if you want no normalization or after-effect, instead pass
* {@link NoNormalization} or {@link NoAfterEffect} respectively.
* @param basicModel Basic model of information content
* @param afterEffect First normalization of information gain
* @param normalization Second (length) normalization
*/
public DFRSimilarity(BasicModel basicModel,
AfterEffect afterEffect,
Normalization normalization) {
if (basicModel == null || afterEffect == null || normalization == null) {
throw new NullPointerException("null parameters not allowed.");
}
this.basicModel = basicModel;
this.afterEffect = afterEffect;
this.normalization = normalization;
}
@Override
protected float score(BasicStats stats, float freq, float docLen) {
float tfn = normalization.tfn(stats, freq, docLen);
return stats.getTotalBoost() *
basicModel.score(stats, tfn) * afterEffect.score(stats, tfn);
}
@Override
protected void explain(Explanation expl,
BasicStats stats, int doc, float freq, float docLen) {
if (stats.getTotalBoost() != 1.0f) {
expl.addDetail(new Explanation(stats.getTotalBoost(), "boost"));
}
Explanation normExpl = normalization.explain(stats, freq, docLen);
float tfn = normExpl.getValue();
expl.addDetail(normExpl);
expl.addDetail(basicModel.explain(stats, tfn));
expl.addDetail(afterEffect.explain(stats, tfn));
}
@Override
public String toString() {
return "DFR " + basicModel.toString() + afterEffect.toString()
+ normalization.toString();
}
/**
* Returns the basic model of information content
*/
public BasicModel getBasicModel() {
return basicModel;
}
/**
* Returns the first normalization
*/
public AfterEffect getAfterEffect() {
return afterEffect;
}
/**
* Returns the second normalization
*/
public Normalization getNormalization() {
return normalization;
}
}