Package index

Rating Data

Defines a common class for rating data and calculates dissimilarities and similarities between ratings by users and items.

ratingMatrix-class ratingMatrix coerce,ratingMatrix,list-method coerce,ratingMatrix,data.frame-method [,ratingMatrix,ANY,ANY,ANY-method sample,ratingMatrix-method colCounts colCounts,ratingMatrix-method rowCounts rowCounts,ratingMatrix-method colMeans,ratingMatrix-method rowMeans,ratingMatrix-method dim,ratingMatrix-method dimnames<-,ratingMatrix,list-method dimnames,ratingMatrix-method nratings nratings,ratingMatrix-method show,ratingMatrix-method image,ratingMatrix-method getNormalize getNormalize,ratingMatrix-method getRatings getRatings,ratingMatrix-method hasRating hasRating,ratingMatrix-method getRatingMatrix getRatingMatrix,ratingMatrix-method

Class "ratingMatrix": Virtual Class for Rating Data

realRatingMatrix-class realRatingMatrix coerce,matrix,realRatingMatrix-method coerce,realRatingMatrix,matrix-method coerce,realRatingMatrix,dgTMatrix-method coerce,dgTMatrix,realRatingMatrix-method coerce,realRatingMatrix,ngCMatrix-method coerce,realRatingMatrix,dgCMatrix-method coerce,dgCMatrix,realRatingMatrix-method coerce,data.frame,realRatingMatrix-method coerce,realRatingMatrix,data.frame-method rowSds rowSds,realRatingMatrix-method colSds colSds,realRatingMatrix-method binarize binarize,realRatingMatrix-method removeKnownRatings removeKnownRatings,realRatingMatrix-method [<-,realRatingMatrix,ANY,ANY,ANY-method getTopNLists getTopNLists,realRatingMatrix-method

Class "realRatingMatrix": Real-valued Rating Matrix

binaryRatingMatrix-class binaryRatingMatrix coerce,matrix,binaryRatingMatrix-method coerce,itemMatrix,binaryRatingMatrix-method coerce,data.frame,binaryRatingMatrix-method coerce,binaryRatingMatrix,matrix-method coerce,binaryRatingMatrix,dgTMatrix-method coerce,binaryRatingMatrix,ngCMatrix-method coerce,binaryRatingMatrix,dgCMatrix-method coerce,binaryRatingMatrix,itemMatrix-method coerce,binaryRatingMatrix,list-method

Class "binaryRatingMatrix": A Binary Rating Matrix

dissimilarity(<binaryRatingMatrix>) dissimilarity(<realRatingMatrix>) similarity()

Dissimilarity and Similarity Calculation Between Rating Data

Data Preparation

Normalizes ratings in real rating matrices and creates list, data frame, and sparse matrix representations.

normalize() denormalize()

Normalize the ratings

getList() getData.frame()

List and Data.frame Representation for Recommender Matrix Objects

dropNA() dropNA2matrix() dropNAis.na()

Sparse Matrix Representation With NAs Not Explicitly Stored

Recommender Models

Learns a recommender model from given data and creates and combines recommendations using recommender algorithms.

Recommender()

Create a Recommender Model

HybridRecommender()

Create a Hybrid Recommender

Recommender-class show,Recommender-method

Class "Recommender": A Recommender Model

Recommendations

Creates recommendations using a recommender model and data about new users; recommendations can be represented as a Top-N list.

predict(<Recommender>)

Predict Recommendations

topNList-class topNList bestN bestN,topNList-method coerce,topNList,dgTMatrix-method coerce,topNList,dgCMatrix-method coerce,topNList,ngCMatrix-method coerce,topNList,matrix-method coerce,topNList,list-method coerce,topNList,realRatingMatrix-method colCounts,topNList-method rowCounts,topNList-method show,topNList-method length,topNList-method removeKnownItems removeKnownItems,topNList-method c,topNList-method

Class "topNList": Top-N List

Evaluation

Creates evaluation schemes from data sets, evaluates recommender models, and calculates prediction accuracy and evaluation metrics.

evaluationScheme()

Creator Function for evaluationScheme

evaluate()

Evaluate a Recommender Models

calcPredictionAccuracy()

Calculate the Prediction Error for a Recommendation

MSE() RMSE() MAE() frobenius()

Error Calculation

plot(<evaluationResults>) plot(<evaluationResultList>)

Plot Evaluation Results

evaluationScheme-class getData getData,evaluationScheme-method show,evaluationScheme-method

Class "evaluationScheme": Evaluation Scheme

evaluationResults-class confusionMatrix-class avg avg,evaluationResults-method getConfusionMatrix getConfusionMatrix,evaluationResults-method getResults getResults,evaluationResults-method getModel getModel,evaluationResults-method getRuns getRuns,evaluationResults-method show,evaluationResults-method

Class "evaluationResults": Results of the Evaluation of a Single Recommender Method

evaluationResultList-class coerce,list,evaluationResultList-method avg,evaluationResultList-method [,evaluationResultList,ANY,missing,missing-method show,evaluationResultList-method

Class "evaluationResultList": Results from Evaluating Multiple Recommender Methods

Datasets

Rating data collected by the Jester Online Joke Recommender System, MovieLens, and the Microsoft web site.

Jester5k

Jester dataset (5k sample)

MSWeb

Anonymous web data from www.microsoft.com

MovieLense

MovieLense Dataset (100k)

Internal Utilities

Utility functions used internally by recommender algorithms.

funkSVD()

Funk SVD for Matrices with Missing Data

returnRatings() getParameters()

Internal Utility Functions