org.lenskit.rerank

Class GreedyRerankingItemRecommender

java.lang.Object

org.lenskit.basic.AbstractItemRecommender

org.lenskit.rerank.GreedyRerankingItemRecommender

All Implemented Interfaces:

ItemRecommender

public class GreedyRerankingItemRecommender
extends AbstractItemRecommender

A hybrid item recommender that uses a greedy re-ranking strategy to allow re-ranking of items iteratively. This general algorithm is commonly employed to efficiently optimize for set properties of a recommendation list such as inter-item diveristy.

This algorithm takes a baseline ranking algorithm, gets the top-n recommendations and re-ranks them iteratively. To select each recommended item, first a scoring algorithm is ran based on the currently selected recommendations and each candidate item. The item with the highest score is then added to the recommended list. This process repeates until enough items are recommended.

Constructor Summary

Constructors

Constructor and Description
GreedyRerankingItemRecommender(ItemRecommender baseRecommender, GreedyRerankStrategy strategy)

Method Summary

Modifier and Type	Method and Description
protected ResultList	recommendWithDetails(long user, int n, LongSet candidateItems, LongSet exclude) Primary method for implementing an item recommender.

Methods inherited from class org.lenskit.basic.AbstractItemRecommender

recommend, recommend, recommend, recommend, recommendWithDetails

Methods inherited from class java.lang.Object

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

Constructor Detail

GreedyRerankingItemRecommender

@Inject
public GreedyRerankingItemRecommender(ItemRecommender baseRecommender,
                                              GreedyRerankStrategy strategy)

Method Detail

recommendWithDetails

protected ResultList recommendWithDetails(long user,
                                          int n,
                                          @Nullable
                                          LongSet candidateItems,
                                          @Nullable
                                          LongSet exclude)

Description copied from class: AbstractItemRecommender

Primary method for implementing an item recommender.

Specified by:

recommendWithDetails in class AbstractItemRecommender

Parameters:

user - The user ID.

n - The number of recommendations to produce, or a negative value to produce unlimited recommendations.

candidateItems - The candidate items, or null for default.

exclude - The exclude set, or null for default.

Returns:

The result list.

See Also:

AbstractItemRecommender.recommendWithDetails(long, int, Set, Set)