Package de.uni_trier.wi2.procake.utils.eval.metrics.k

Class KEvalMetric

java.lang.Object

de.uni_trier.wi2.procake.utils.eval.metrics.k.KEvalMetric

All Implemented Interfaces:

EvalMetric

Direct Known Subclasses:

HitsMetric, QualityMuellerMetric, QualityStromerMetric

public abstract class KEvalMetric
extends Object
implements EvalMetric

Interface for metrics that only look at the top-k cases in the ranking.

Author:

Maximilian Hoffmann

Constructor Summary

Constructors

Constructor	Description
KEvalMetric()
KEvalMetric(int k)

Method Summary

Modifier and Type	Method	Description
abstract double	computeEvalKMetric(SimpleSimilarityResult groundTruthRankingResult, SimpleSimilarityResult predictedRankingResult, Integer k)	Computes a ranking metric for the top-k cases of the ranking results.
double	computeEvalMetric(SimpleSimilarityResult groundTruthResult, SimpleSimilarityResult predictedResult)	Computes a ranking metric for the complete ranking.
int	getK()
String	getMetricName()	Returns a name for the metric to be identified in any string representations.
void	setK(int k)

Methods inherited from class java.lang.Object

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

Constructor Details

KEvalMetric

public KEvalMetric()

KEvalMetric

public KEvalMetric(int k)

Method Details

computeEvalMetric

public double computeEvalMetric(SimpleSimilarityResult groundTruthResult,
 SimpleSimilarityResult predictedResult)

Description copied from interface: EvalMetric

Computes a ranking metric for the complete ranking.

Specified by:

computeEvalMetric in interface EvalMetric

Parameters:

groundTruthResult - the ground-truth ranking

predictedResult - the predicted ranking (to analyze)

Returns:

a double score

getMetricName

public String getMetricName()

Description copied from interface: EvalMetric

Returns a name for the metric to be identified in any string representations.

Specified by:

getMetricName in interface EvalMetric

Returns:

the name of the metric

getK

public int getK()

setK

public void setK(int k)

computeEvalKMetric

public abstract double computeEvalKMetric(SimpleSimilarityResult groundTruthRankingResult,
 SimpleSimilarityResult predictedRankingResult,
 Integer k)

Computes a ranking metric for the top-k cases of the ranking results.

Parameters:

groundTruthRankingResult - the ground-truth ranking

predictedRankingResult - the predicted ranking results to analyze

k - the number of top cases to look at

Returns:

a score for the ranking according to this metric