de.lmu.ifi.dbs.elki.evaluation.scores

Class DCGEvaluation

java.lang.Object

de.lmu.ifi.dbs.elki.evaluation.scores.DCGEvaluation

All Implemented Interfaces:

ScoreEvaluation

@Reference(authors="K. J\u00e4rvelin, J. Kek\u00e4l\u00e4inen",
           title="Cumulated gain-based evaluation of IR techniques",
           booktitle="ACM Transactions on Information Systems (TOIS)",
           url="https://doi.org/10.1145/582415.582418",
           bibkey="DBLP:journals/tois/JarvelinK02")
public class DCGEvaluation
extends java.lang.Object
implements ScoreEvaluation

Discounted Cumulative Gain.

This evaluation metric would be able to use relevance information, but the current implementation is for binary labels only (it is easy to add, but requires API additions or changes).

Reference:

K. Järvelin, J. Kekäläinen
Cumulated gain-based evaluation of IR techniques
ACM Transactions on Information Systems (TOIS)

TODO: support weighted ground truth.

TODO: allow using other logarithms than 2?

Since:

0.7.5

Author:

Erich Schubert

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	DCGEvaluation.Parameterizer Parameterization class.

Nested classes/interfaces inherited from interface de.lmu.ifi.dbs.elki.evaluation.scores.ScoreEvaluation

ScoreEvaluation.Predicate<T>, ScoreEvaluation.ScoreIter

Field Summary

Fields

Modifier and Type	Field and Description
static DCGEvaluation	STATIC Static instance

Constructor Summary

Constructors

Constructor and Description
DCGEvaluation()

Method Summary

Modifier and Type	Method and Description
static <I extends ScoreEvaluation.ScoreIter> double	computeDCG(ScoreEvaluation.Predicate<? super I> predicate, I iter) Compute the DCG given a set of positive IDs and a sorted list of entries, which may include ties.
<I extends ScoreEvaluation.ScoreIter> double	evaluate(ScoreEvaluation.Predicate<? super I> predicate, I iter) Evaluate a given predicate and iterator.
double	expected(int pos, int all) Expected score for a random result.
static double	maximum(int pos) Maximum DCG.
static double	sumInvLog1p(int s, int e) Compute \sum_{i=s}^e 1/log(1+i)

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.lmu.ifi.dbs.elki.evaluation.scores.ScoreEvaluation

evaluate

Field Detail

STATIC

public static final DCGEvaluation STATIC

Static instance

Constructor Detail

DCGEvaluation

public DCGEvaluation()

Method Detail

evaluate

public <I extends ScoreEvaluation.ScoreIter> double evaluate(ScoreEvaluation.Predicate<? super I> predicate,
                                                             I iter)

Description copied from interface: ScoreEvaluation

Evaluate a given predicate and iterator.

Specified by:

evaluate in interface ScoreEvaluation

Type Parameters:

I - Iterator type

Parameters:

predicate - Predicate (for positives)

iter - Iterator

Returns:

Score

expected

public double expected(int pos,
                       int all)

Description copied from interface: ScoreEvaluation

Expected score for a random result.

Specified by:

expected in interface ScoreEvaluation

Parameters:

pos - Number of positive elements

all - Total number of elements

Returns:

Expected score

sumInvLog1p

public static double sumInvLog1p(int s,
                                 int e)

Compute \sum_{i=s}^e 1/log(1+i)

Parameters:

s - Start value

e - End value (inclusive!)

Returns:

Sum

computeDCG

public static <I extends ScoreEvaluation.ScoreIter> double computeDCG(ScoreEvaluation.Predicate<? super I> predicate,
                                                                      I iter)

Compute the DCG given a set of positive IDs and a sorted list of entries, which may include ties.

Type Parameters:

I - Iterator type

Parameters:

predicate - Predicate to test for positive objects

iter - Iterator over results, with ties.

Returns:

area under curve

maximum

public static double maximum(int pos)

Maximum DCG.

Parameters:

pos - Number of positive objects

Returns:

Max