org.grouplens.lenskit.knn.user

Class LiveNeighborFinder

java.lang.Object

org.grouplens.lenskit.knn.user.LiveNeighborFinder

All Implemented Interfaces:

NeighborFinder

public class LiveNeighborFinder
extends Object
implements NeighborFinder

Neighborhood finder that does a fresh search over the data source ever time.

Constructor Summary

Constructors

Constructor and Description
LiveNeighborFinder(UserEventDAO udao, ItemEventDAO idao, UserSimilarity sim, UserVectorNormalizer norm, Threshold thresh) Construct a new user neighborhood finder.

Method Summary

Modifier and Type	Method and Description
Iterable<Neighbor>	getCandidateNeighbors(UserHistory<? extends Event> user, LongSet items) Get potential neighbors for a particular user.

Methods inherited from class java.lang.Object

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

Constructor Detail

LiveNeighborFinder

@Inject
public LiveNeighborFinder(UserEventDAO udao,
                                  ItemEventDAO idao,
                                  UserSimilarity sim,
                                  UserVectorNormalizer norm,
                                  @UserSimilarityThreshold
                                  Threshold thresh)

Construct a new user neighborhood finder.

Parameters:

udao - The user-event DAO.

idao - The item-event DAO.

sim - The similarity function to use.

norm - The normalizer for user rating/preference vectors.

thresh - The threshold for user similarities.

Method Detail

getCandidateNeighbors

public Iterable<Neighbor> getCandidateNeighbors(UserHistory<? extends Event> user,
                                                LongSet items)

Description copied from interface: NeighborFinder

Get potential neighbors for a particular user.

Specified by:

getCandidateNeighbors in interface NeighborFinder

Parameters:

user - The user whose neighbors are wanted.

items - The items that the client needs to be able to score or recommend.

Returns:

A collection of potential neighbors for user. This collection may include neighbors that are not useful for scoring any item in items; the item set is just to help the neighbor finder guide its search if relevant.