Recommender: Create a Recommender Model

In recommenderlab: Lab for Developing and Testing Recommender Algorithms

Create a Recommender Model

Description

Learns a recommender model from given data.

Usage

Recommender(data, ...)
## S4 method for signature 'ratingMatrix'
Recommender(data, method, parameter=NULL)

Arguments

data	training data.
method	a character string defining the recommender method to use (see details).
parameter	parameters for the recommender algorithm.
...	further arguments.

Details

Recommender uses the registry mechanism from package registry to manage methods. This lets users easily specify and add new methods. The registry is called recommenderRegistry. See examples section. Methods SVD and LIBMF require the suggested packages irlba and recosystem, respectively. Install the corresponding package before using either method.

Value

An object of class 'Recommender'.

See Also

Recommender, ratingMatrix, predict.

Other recommender models: HybridRecommender(), Recommender-class

Examples

data("MSWeb")
MSWeb10 <- sample(MSWeb[rowCounts(MSWeb) >10,], 100)

rec <- Recommender(MSWeb10, method = "POPULAR")
rec

getModel(rec)

## save and read a recommender model
saveRDS(rec, file = "rec.rds")
rec2 <- readRDS("rec.rds")
rec2
unlink("rec.rds")

## look at registry and a few methods
recommenderRegistry$get_entry_names()

recommenderRegistry$get_entry("POPULAR", dataType = "binaryRatingMatrix")

recommenderRegistry$get_entry("SVD", dataType = "realRatingMatrix")

recommenderlab documentation built on Oct. 5, 2026, 9:08 a.m.