org.elasticsearch.search.aggregations.bucket.terms.heuristic.MutualInformation Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of elasticsearch Show documentation
Show all versions of elasticsearch Show documentation
Elasticsearch subproject :server
/*
* Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
* or more contributor license agreements. Licensed under the Elastic License
* 2.0 and the Server Side Public License, v 1; you may not use this file except
* in compliance with, at your election, the Elastic License 2.0 or the Server
* Side Public License, v 1.
*/
package org.elasticsearch.search.aggregations.bucket.terms.heuristic;
import org.elasticsearch.common.io.stream.StreamInput;
import org.elasticsearch.xcontent.ConstructingObjectParser;
import org.elasticsearch.xcontent.XContentBuilder;
import java.io.IOException;
public class MutualInformation extends NXYSignificanceHeuristic {
public static final String NAME = "mutual_information";
public static final ConstructingObjectParser PARSER = new ConstructingObjectParser<>(
NAME,
buildFromParsedArgs(MutualInformation::new)
);
static {
NXYSignificanceHeuristic.declareParseFields(PARSER);
}
private static final double log2 = Math.log(2.0);
public MutualInformation(boolean includeNegatives, boolean backgroundIsSuperset) {
super(includeNegatives, backgroundIsSuperset);
}
/**
* Read from a stream.
*/
public MutualInformation(StreamInput in) throws IOException {
super(in);
}
@Override
public boolean equals(Object other) {
if ((other instanceof MutualInformation) == false) {
return false;
}
return super.equals(other);
}
@Override
public int hashCode() {
int result = NAME.hashCode();
result = 31 * result + super.hashCode();
return result;
}
/**
* Calculates mutual information
* see "Information Retrieval", Manning et al., Eq. 13.17
*/
@Override
public double getScore(long subsetFreq, long subsetSize, long supersetFreq, long supersetSize) {
Frequencies frequencies = computeNxys(subsetFreq, subsetSize, supersetFreq, supersetSize, "MutualInformation");
double score = (getMITerm(frequencies.N00, frequencies.N0_, frequencies.N_0, frequencies.N) + getMITerm(
frequencies.N01,
frequencies.N0_,
frequencies.N_1,
frequencies.N
) + getMITerm(frequencies.N10, frequencies.N1_, frequencies.N_0, frequencies.N) + getMITerm(
frequencies.N11,
frequencies.N1_,
frequencies.N_1,
frequencies.N
)) / log2;
if (Double.isNaN(score)) {
score = Double.NEGATIVE_INFINITY;
}
// here we check if the term appears more often in subset than in background without subset.
if (includeNegatives == false && frequencies.N11 / frequencies.N_1 < frequencies.N10 / frequencies.N_0) {
score = Double.NEGATIVE_INFINITY;
}
return score;
}
/* make sure that
0 * log(0/0) = 0
0 * log(0) = 0
Else, this would be the score:
double score =
N11 / N * Math.log((N * N11) / (N1_ * N_1))
+ N01 / N * Math.log((N * N01) / (N0_ * N_1))
+ N10 / N * Math.log((N * N10) / (N1_ * N_0))
+ N00 / N * Math.log((N * N00) / (N0_ * N_0));
but we get many NaN if we do not take case of the 0s */
static double getMITerm(double Nxy, double Nx_, double N_y, double N) {
double numerator = Math.abs(N * Nxy);
double denominator = Math.abs(Nx_ * N_y);
double factor = Math.abs(Nxy / N);
if (numerator < 1.e-7 && factor < 1.e-7) {
return 0.0;
} else {
return factor * Math.log(numerator / denominator);
}
}
@Override
public String getWriteableName() {
return NAME;
}
@Override
public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {
builder.startObject(NAME);
super.build(builder);
builder.endObject();
return builder;
}
public static class MutualInformationBuilder extends NXYBuilder {
public MutualInformationBuilder(boolean includeNegatives, boolean backgroundIsSuperset) {
super(includeNegatives, backgroundIsSuperset);
}
@Override
public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {
builder.startObject(NAME);
super.build(builder);
builder.endObject();
return builder;
}
}
}