es.uam.eps.ir.ranksys.rec.runner.fast

Class FastFilters

java.lang.Object

es.uam.eps.ir.ranksys.rec.runner.fast.FastFilters

public class FastFilters
extends Object

Filters for the filter recommender method in FastRecommender.

Author:

Saúl Vargas (saul.vargas@uam.es)

Constructor Summary

Constructors

Constructor and Description
FastFilters()

Method Summary

Modifier and Type	Method and Description
static <U,I> Function<U,IntPredicate>	all() False item filter as it admits every item.
static <U> Function<U,IntPredicate>	and(Function<U,IntPredicate>... filters) AND of two or more filters.
static <U,I> Function<U,IntPredicate>	notInTrain(FastPreferenceData<U,I> trainData) Item filter that discards items in the training preference data.
static <U> Function<U,IntPredicate>	notSelf(FastUserIndex<U> users) For social network recommendations, void a user being recommended to herself.
static <U,I,F> Function<U,IntPredicate>	withFeatures(FastFeatureData<I,F,?> featureData) Item filter that discard items for which no feature data is available.

Methods inherited from class java.lang.Object

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

Constructor Detail

FastFilters

public FastFilters()

Method Detail

all

public static <U,I> Function<U,IntPredicate> all()

False item filter as it admits every item.

Type Parameters:

U - type of the users

I - type of the items

Returns:

item filters that return true always

notInTrain

public static <U,I> Function<U,IntPredicate> notInTrain(FastPreferenceData<U,I> trainData)

Item filter that discards items in the training preference data.

Type Parameters:

U - type of the users

I - type of the items

Parameters:

trainData - preference data

Returns:

item filters for each using returning true if the user-item pair was not observed in the preference data

withFeatures

public static <U,I,F> Function<U,IntPredicate> withFeatures(FastFeatureData<I,F,?> featureData)

Item filter that discard items for which no feature data is available.

Type Parameters:

U - type of the users

I - type of the items

F - type of the features

Parameters:

featureData - feature data

Returns:

item filters that return true when there is any feature information for the item

notSelf

public static <U> Function<U,IntPredicate> notSelf(FastUserIndex<U> users)

For social network recommendations, void a user being recommended to herself.

Type Parameters:

U - type of the user

Parameters:

users - user index

Returns:

item-user filter that return true if the recommended item-user is not the target user.

and

public static <U> Function<U,IntPredicate> and(Function<U,IntPredicate>... filters)

AND of two or more filters.

Type Parameters:

U - type of the users

Parameters:

filters - a number of item filters

Returns:

an item filter which does a logical AND to two or more filters