Package | Description |
---|---|
es.uam.eps.ir.ranksys.fast.index |
Fast versions of user/item/feature indexes.
|
es.uam.eps.ir.ranksys.fast.preference |
Fast versions of PreferenceData.
|
es.uam.eps.ir.ranksys.mf |
Matrix factorization base classes.
|
es.uam.eps.ir.ranksys.mf.rec |
Matrix factorization recommenders.
|
es.uam.eps.ir.ranksys.nn.item |
Item-based nearest neighbors recommenders.
|
es.uam.eps.ir.ranksys.nn.user |
User-based nearest neighbors recommenders.
|
es.uam.eps.ir.ranksys.nn.user.neighborhood |
User neighborhoods.
|
es.uam.eps.ir.ranksys.nn.user.sim |
User similarities.
|
es.uam.eps.ir.ranksys.novdiv.reranking |
Re-rankers.
|
es.uam.eps.ir.ranksys.novelty.inverted.neighborhood |
Inverted neighborhoods.
|
es.uam.eps.ir.ranksys.rec.fast |
Fast version of base classes and interfaces for recommendation algorithms.
|
es.uam.eps.ir.ranksys.rec.fast.basic |
Basic recommendation algorithms.
|
es.uam.eps.ir.ranksys.rec.runner.fast |
Fast recommender runners.
|
org.ranksys.compression.preferences |
Compressed PreferenceData for binary and rating data.
|
org.ranksys.fast.preference |
Fast versions of PreferenceData.
|
org.ranksys.fm |
Wrappers for factorisation machines.
|
org.ranksys.fm.rec |
FM-based recommenders.
|
org.ranksys.formats.factorization |
Write and read format for factorised models.
|
org.ranksys.lda |
LDA topic modelling and recommender.
|
org.ranksys.rec.fast |
Fast version of base classes and interfaces for recommendation algorithms.
|
Modifier and Type | Class and Description |
---|---|
class |
SimpleFastUserIndex<U>
Simple implementation of FastUserIndex backed by a bi-map IdxIndex
|
Modifier and Type | Interface and Description |
---|---|
interface |
FastPreferenceData<U,I>
Fast version of PreferenceData, where preferences for users and items are
stored internally by the indexes provided by FastUserIndex and FastItemIndex.
|
Modifier and Type | Class and Description |
---|---|
class |
AbstractFastPreferenceData<U,I>
Abstract FastFeatureData, implementing the interfaces of FastUserIndex and FastItemIndex by delegating to implementations of these.
|
class |
SimpleFastPreferenceData<U,I>
Simple implementation of FastPreferenceData backed by nested lists.
|
class |
TransposedPreferenceData<I,U>
Transposed preferences, where users and items change roles.
|
Modifier and Type | Field and Description |
---|---|
protected FastUserIndex<U> |
AbstractFastPreferenceData.ui
user index.
|
Modifier and Type | Method and Description |
---|---|
static <U,I> SimpleFastPreferenceData<U,I> |
SimpleFastPreferenceData.load(Stream<org.jooq.lambda.tuple.Tuple3<U,I,Double>> tuples,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Loads a SimpleFastPreferenceData from a stream of user-item-value triples.
|
static <U,I,O> SimpleFastPreferenceData<U,I> |
SimpleFastPreferenceData.load(Stream<org.jooq.lambda.tuple.Tuple4<U,I,Double,O>> tuples,
org.jooq.lambda.function.Function4<Integer,Integer,Double,O,? extends IdxPref> uIdxPrefFun,
org.jooq.lambda.function.Function4<Integer,Integer,Double,O,? extends IdxPref> iIdxPrefFun,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
Function<IdxPref,IdPref<I>> uIdPrefFun,
Function<IdxPref,IdPref<U>> iIdPrefFun)
Loads a SimpleFastPreferenceData from a stream of user-item-value-other tuples.
|
Constructor and Description |
---|
AbstractFastPreferenceData(FastUserIndex<U> users,
FastItemIndex<I> items)
Constructor.
|
AbstractFastPreferenceData(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor.
|
SimpleFastPreferenceData(int numPreferences,
List<List<IdxPref>> uidxList,
List<List<IdxPref>> iidxList,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Constructor with default IdxPref to IdPref converter.
|
SimpleFastPreferenceData(int numPreferences,
List<List<IdxPref>> uidxList,
List<List<IdxPref>> iidxList,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor with custom IdxPref to IdPref converter.
|
Modifier and Type | Class and Description |
---|---|
class |
Factorization<U,I>
Matrix factorization.
|
Modifier and Type | Field and Description |
---|---|
protected FastUserIndex<U> |
Factorization.uIndex
user index
|
Constructor and Description |
---|
Factorization(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
cern.colt.matrix.impl.DenseDoubleMatrix2D userMatrix,
cern.colt.matrix.impl.DenseDoubleMatrix2D itemMatrix,
int K)
Constructor for stored factorizations.
|
Factorization(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
int K,
cern.colt.function.DoubleFunction initFunction)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
MFRecommender<U,I>
Matrix factorization recommender.
|
Constructor and Description |
---|
MFRecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
Factorization<U,I> factorization)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
ItemNeighborhoodRecommender<U,I>
Item-based nearest neighbors recommender.
|
Modifier and Type | Class and Description |
---|---|
class |
UserNeighborhoodRecommender<U,I>
User-based nearest neighbors recommender.
|
Modifier and Type | Class and Description |
---|---|
class |
CachedUserNeighborhood<U>
Cached user similarity.
|
class |
ThresholdUserNeighborhood<U>
Threshold user neighborhood.
|
class |
TopKUserNeighborhood<U>
User top-K neighborhood.
|
class |
UserNeighborhood<U>
User neighborhood.
|
Modifier and Type | Field and Description |
---|---|
protected FastUserIndex<U> |
UserNeighborhood.uIndex
Fast user index.
|
Constructor and Description |
---|
CachedUserNeighborhood(FastUserIndex<U> uIndex,
Stream<org.jooq.lambda.tuple.Tuple2<U,Stream<Tuple2od<U>>>> neighborhoods)
Constructor that caches a stream of previously calculated neighborhoods.
|
UserNeighborhood(FastUserIndex<U> uIndex,
Neighborhood neighborhood)
Constructor
|
Modifier and Type | Class and Description |
---|---|
class |
SetCosineUserSimilarity<U>
Set cosine user similarity.
|
class |
SetJaccardUserSimilarity<U>
Set Jaccard user similarity.
|
class |
UserSimilarity<U>
User similarity.
|
class |
VectorCosineUserSimilarity<U>
Vector cosine user similarity.
|
class |
VectorJaccardUserSimilarity<U>
Vector Jaccard user similarity.
|
Modifier and Type | Field and Description |
---|---|
protected FastUserIndex<U> |
UserSimilarity.uIndex
Fast user index.
|
Constructor and Description |
---|
UserSimilarity(FastUserIndex<U> uIndex,
Similarity sim)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
RerankingRecommender<U,I>
Wrapper for re-ranker that re-ranks the output of another recommender.
|
Constructor and Description |
---|
RerankingRecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
FastRecommender<U,I> recommender,
Reranker<U,I> reranker)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
InvertedUserNeighborhood<U>
Inverted user neighborhood.
|
Modifier and Type | Interface and Description |
---|---|
interface |
FastRecommender<U,I>
Fast recommender.
|
Modifier and Type | Class and Description |
---|---|
class |
AbstractFastRecommender<U,I>
Abstract (fast) recommender.
|
class |
FastRankingRecommender<U,I>
Recommender for top-n recommendations.
|
Modifier and Type | Field and Description |
---|---|
protected FastUserIndex<U> |
AbstractFastRecommender.uIndex
Fast user index.
|
Constructor and Description |
---|
AbstractFastRecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Constructor.
|
FastRankingRecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
PopularityRecommender<U,I>
Popularity-based recommender.
|
class |
RandomRecommender<U,I>
Random recommender.
|
Constructor and Description |
---|
RandomRecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Constructor.
|
Modifier and Type | Method and Description |
---|---|
static <U> Function<U,IntPredicate> |
FastFilters.notSelf(FastUserIndex<U> users)
For social network recommendations, void a user being recommended to herself.
|
Constructor and Description |
---|
FastFilterRecommenderRunner(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex,
Stream<U> users,
Function<U,IntPredicate> userFilter,
int maxLength)
Constructor.
|
Modifier and Type | Class and Description |
---|---|
class |
AbstractCODECPreferenceData<U,I,Cu,Ci>
Abstract PreferenceData using compression.
|
class |
BinaryCODECPreferenceData<U,I,Cu,Ci>
PreferenceData for binary data using compression.
|
class |
RatingCODECPreferenceData<U,I,Cu,Ci,Cv>
PreferenceData for rating data using compression.
|
Constructor and Description |
---|
AbstractCODECPreferenceData(FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec)
Constructor with default IdxPref to IdPref converters.
|
AbstractCODECPreferenceData(FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor with custom IdxPref to IdPref converters.
|
BinaryCODECPreferenceData(FastPreferenceData<U,I> preferences,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec)
Constructor that utilizes other PreferenceData object with default IdxPref to IdPref
converters.
|
BinaryCODECPreferenceData(FastPreferenceData<U,I> preferences,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor that utilizes other PreferenceData object with custom IdxPref to IdPref converters.
|
BinaryCODECPreferenceData(Stream<Tuple2io<int[]>> ul,
Stream<Tuple2io<int[]>> il,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec)
Constructor using streams of user and items preferences lists with default IdxPref to IdPref
converters.
|
BinaryCODECPreferenceData(Stream<Tuple2io<int[]>> ul,
Stream<Tuple2io<int[]>> il,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor using streams of user and items preferences lists with custom IdxPref to IdPref converters.
|
RatingCODECPreferenceData(FastPreferenceData<U,I> preferences,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
CODEC<Cv> r_codec)
Constructor that utilizes other PreferenceData object with default IdxPref to IdPref
converters.
|
RatingCODECPreferenceData(FastPreferenceData<U,I> preferences,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
CODEC<Cv> r_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor that utilizes other PreferenceData object with custom IdxPref to IdPref
converters.
|
RatingCODECPreferenceData(FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
CODEC<Cv> r_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor custom IdxPref to IdPref converters that does not read any input data.
|
RatingCODECPreferenceData(Stream<Tuple2io<int[][]>> ul,
Stream<Tuple2io<int[][]>> il,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
CODEC<Cv> r_codec)
Constructor using streams of user and items preferences lists with default IdxPref to IdPref
converters.
|
RatingCODECPreferenceData(Stream<Tuple2io<int[][]>> ul,
Stream<Tuple2io<int[][]>> il,
FastUserIndex<U> users,
FastItemIndex<I> items,
CODEC<Cu> u_codec,
CODEC<Ci> i_codec,
CODEC<Cv> r_codec,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor using streams of user and items preferences lists with custom IdxPref to IdPref
converters.
|
Modifier and Type | Interface and Description |
---|---|
interface |
FastPointWisePreferenceData<U,I>
Fast point-wise preference data.
|
interface |
FastPreferenceSampler<U,I>
Fast preference sampler, for stochastic algorithms.
|
Modifier and Type | Class and Description |
---|---|
class |
AbstractFastPreferenceSampler<U,I>
Abstract fast preference sampler, implementing methods from
FastUserIndex and FastItemIndex.
|
class |
IteratorsAbstractFastPreferenceData<U,I>
Extends AbstractFastPreferenceData and implements the data access stream-based methods using the iterator-based ones.
|
class |
StreamsAbstractFastPreferenceData<U,I>
Extends AbstractFastPreferenceData and implements the data access iterator-based methods
using the stream-based ones.
|
Constructor and Description |
---|
AbstractFastPreferenceSampler(FastUserIndex<U> users,
FastItemIndex<I> items)
Constructor.
|
IteratorsAbstractFastPreferenceData(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex)
Constructor with default IdxPref to IdPref converter.
|
IteratorsAbstractFastPreferenceData(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor with custom IdxPref to IdPref converter.
|
StreamsAbstractFastPreferenceData(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex)
Constructor with default IdxPref to IdPref converter.
|
StreamsAbstractFastPreferenceData(FastUserIndex<U> userIndex,
FastItemIndex<I> itemIndex,
Function<IdxPref,IdPref<I>> uPrefFun,
Function<IdxPref,IdPref<U>> iPrefFun)
Constructor with custom IdxPref to IdPref converter.
|
Modifier and Type | Class and Description |
---|---|
class |
PreferenceFM<U,I>
Wraps a factorisation machine to work with RankSys user-preference pairs.
|
Constructor and Description |
---|
PreferenceFM(FastUserIndex<U> users,
FastItemIndex<I> items,
org.ranksys.javafm.FM fm)
Constructor with default converter to IdxPref.
|
PreferenceFM(FastUserIndex<U> users,
FastItemIndex<I> items,
org.ranksys.javafm.FM fm,
Function<IdPref<I>,IdxPref> uPrefFun)
Constructor with custom default converter to IdxPref.
|
Modifier and Type | Class and Description |
---|---|
class |
FMRecommender<U,I>
A recommender using a factorisation machine.
|
Modifier and Type | Method and Description |
---|---|
<U,I> PreferenceFM |
FMFormat.load(InputStream in,
FastUserIndex<U> users,
FastItemIndex<I> items)
Loads a factorisation machine from an input stream.
|
<U,I> Factorization<U,I> |
SimpleFactorizationFormat.load(InputStream in,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex) |
<U,I> PreferenceFM<U,I> |
SimpleFMFormat.load(InputStream in,
FastUserIndex<U> users,
FastItemIndex<I> items) |
<U,I> Factorization<U,I> |
FactorizationFormat.load(InputStream in,
FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex)
Loads a matrix from a compressed input stream.
|
Modifier and Type | Class and Description |
---|---|
class |
LDARecommender<U,I>
LDA recommender.
|
Constructor and Description |
---|
LDARecommender(FastUserIndex<U> uIndex,
FastItemIndex<I> iIndex,
cc.mallet.topics.ParallelTopicModel topicModel)
Constructor
|
Modifier and Type | Class and Description |
---|---|
class |
FastEnsembleRecommender<U,I>
Ensemble of recommenders, performs a linear combination of the scores
given by several recommenders.
|
Copyright © 2016. All rights reserved.