Package es.uam.eps.ir.relison.links.recommendation.updateable.mf

Class UpdateableFactorizer<U,I>

java.lang.Object

es.uam.eps.ir.relison.links.recommendation.updateable.mf.UpdateableFactorizer<U,I>

Type Parameters:

U - type of the users

I - type of the items

Direct Known Subclasses:

ALSUpdateableFactorizer

public abstract class UpdateableFactorizer<U,I>
extends java.lang.Object

UpdateableFactorizer. Abstract class for matrix factorization algorithms.

Constructor Summary

Constructors

Constructor	Description
UpdateableFactorizer()

Method Summary

Modifier and Type	Method	Description
abstract double	error(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data)	Global loss of the factorization.
abstract UpdateableFactorization<U,I>	factorize(int K, FastUpdateablePreferenceData<U,I> data)	Creates and calculates a factorization.
abstract void	factorize(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data)	Calculates the factorization by using a previously generate matrix factorization.
abstract void	update(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data, U u, I i, double weight)	Updates the factorization.
abstract void	updateDelete(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data, U u, I i)	Updates the factorization when a rating is removed.

Methods inherited from class java.lang.Object

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

Constructor Details

UpdateableFactorizer

public UpdateableFactorizer()

Method Details

error

public abstract double error(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data)

Global loss of the factorization.

Parameters:

factorization - matrix factorization

data - preference data

Returns:

the global loss

factorize

public abstract UpdateableFactorization<U,I> factorize(int K, FastUpdateablePreferenceData<U,I> data)

Creates and calculates a factorization.

Parameters:

K - size of the latent feature space.

data - preference data

Returns:

a matrix factorization

factorize

public abstract void factorize(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data)

Calculates the factorization by using a previously generate matrix factorization.

Parameters:

factorization - matrix factorization

data - preference data

update

public abstract void update(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data, U u, I i, double weight)

Updates the factorization.

Parameters:

factorization - an updateable factorization

data - preference data

u - the user

i - the item

weight - the rating for the item i provided by the user u

updateDelete

public abstract void updateDelete(UpdateableFactorization<U,I> factorization, FastUpdateablePreferenceData<U,I> data, U u, I i)

Updates the factorization when a rating is removed.

Parameters:

factorization - an updateable factorization

data - preference data

u - the user

i - the item