es.uam.eps.ir.ranksys.diversity.sales.metrics

Class AbstractSalesDiversityMetric<U,I>

java.lang.Object

es.uam.eps.ir.ranksys.metrics.AbstractSystemMetric<U,I>

es.uam.eps.ir.ranksys.diversity.sales.metrics.AbstractSalesDiversityMetric<U,I>

Type Parameters:

U - type of the users

I - type of the items

All Implemented Interfaces:

SystemMetric<U,I>

Direct Known Subclasses:

EIUDC, EIUFD, EIURD, GiniIndex

public abstract class AbstractSalesDiversityMetric<U,I>
extends AbstractSystemMetric<U,I>

Abstract sales diversity metrics. It handles the counting of how many times an item is recommended.

Author:

Saúl Vargas (saul.vargas@uam.es)

Field Summary

Fields

Modifier and Type	Field and Description
protected int	cutoff maximum length of the recommendation lists that is evaluated
protected double	freeNorm Norm for free discovery item novelty models.
protected it.unimi.dsi.fastutil.objects.Object2DoubleOpenHashMap<I>	itemCount Map of item-count.
protected it.unimi.dsi.fastutil.objects.Object2DoubleOpenHashMap<I>	itemWeight Map of item-weight.
protected int	numUsers Number of users.

Constructor Summary

Constructors

Constructor and Description
AbstractSalesDiversityMetric(int cutoff, RankingDiscountModel disc, RelevanceModel<U,I> rel) Constructor

Method Summary

Modifier and Type	Method and Description
void	add(Recommendation<U,I> recommendation)
void	combine(SystemMetric<U,I> other)
double	evaluate()
protected abstract double	nov(I i) Returns the sales novelty of an item.
void	reset()

Methods inherited from class java.lang.Object

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

Field Detail

cutoff

protected final int cutoff

maximum length of the recommendation lists that is evaluated

itemCount

protected final it.unimi.dsi.fastutil.objects.Object2DoubleOpenHashMap<I> itemCount

Map of item-count.

itemWeight

protected final it.unimi.dsi.fastutil.objects.Object2DoubleOpenHashMap<I> itemWeight

Map of item-weight.

freeNorm

protected double freeNorm

Norm for free discovery item novelty models.

numUsers

protected int numUsers

Number of users.

Constructor Detail

AbstractSalesDiversityMetric

public AbstractSalesDiversityMetric(int cutoff,
                                    RankingDiscountModel disc,
                                    RelevanceModel<U,I> rel)

Constructor

Parameters:

cutoff - maximum length of the recommendation lists that is evaluated

disc - ranking discount model

rel - relevance model

Method Detail

add

public void add(Recommendation<U,I> recommendation)

combine

public void combine(SystemMetric<U,I> other)

nov

protected abstract double nov(I i)

Returns the sales novelty of an item.

Parameters:

i - item

Returns:

the sales novelty of the item

evaluate

public double evaluate()

reset

public void reset()