es.upm.etsisi.cf4j.recommender

Class Recommender

java.lang.Object

es.upm.etsisi.cf4j.recommender.Recommender

Direct Known Subclasses:

BiasedMF, BNMF, CLiMF, HPF, ItemKNN, NMF, PMF, SVDPlusPlus, URP, UserKNN

public abstract class Recommender
extends Object

Abstract class that represents any recommender. To build a new recommender following methods must be implemented:

• fit(): fits the recommender according to an instance of a DataModel.
• predict(userIndex, itemIndex): computes a rating prediction for a given user and item.

Field Summary

Fields

Modifier and Type	Field and Description
protected DataModel	datamodel DataModel instance used for the Recommender

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	Recommender(DataModel datamodel) Recommender constructor

Method Summary

Modifier and Type	Method and Description
abstract void	fit() Estimates model parameters given the hyper-parameters
DataModel	getDataModel() Returns the DataModel instance
abstract double	predict(int userIndex, int itemIndex) Computes a rating prediction
double[]	predict(TestUser testUser) Computes the rating predictions of the TestItems rated by a TestUser

Methods inherited from class java.lang.Object

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

Field Detail

datamodel

protected DataModel datamodel

DataModel instance used for the Recommender

Constructor Detail

Recommender

protected Recommender(DataModel datamodel)

Recommender constructor

Parameters:

datamodel - instance of a DataModel

Method Detail

getDataModel

public DataModel getDataModel()

Returns the DataModel instance

Returns:

DataModel instance used by the Recommender

fit

public abstract void fit()

Estimates model parameters given the hyper-parameters

predict

public abstract double predict(int userIndex,
                               int itemIndex)

Computes a rating prediction

Parameters:

userIndex - Index of the user in the array of Users of the DataModel instance

itemIndex - Index of the item in the array of Items of the DataModel instance

Returns:

Prediction

predict

public double[] predict(TestUser testUser)

Computes the rating predictions of the TestItems rated by a TestUser

Parameters:

testUser - TestUser for which to calculate resting predictions

Returns:

Rating prediction for TestItems rated by testUser. Positions of this array overlaps with testItemIndexes returned by testUser.getTestItemAt(pos)