There is a distributed recommender that uses matrix factorization via Alternating Least Squares. Due to Hadoop's inefficiency in executing iterative algorithms, its very slow unfortunately.
The examples module contains code to run this recommender on the movielens and netflix datasets. --sebastian Am 04.04.2012 16:56 schrieb "Stemmer, Maya" <[email protected]>: > Hi, > > "Mahout in Action" refers to the 0.5 release of Mahout. > I'm using the 0.6 release, and I wanted to know where I can find a more > up-to-date documentation of the available implementations. > For example, the book mentions that the current implementation of the SVD > recommender computes the SVD in memory. > Is there a newer implementation - one that can be distributed? > > Thanks, > Maya > > --------------------------------------------------------------------- > Intel Electronics Ltd. > > This e-mail and any attachments may contain confidential material for > the sole use of the intended recipient(s). Any review or distribution > by others is strictly prohibited. If you are not the intended > recipient, please contact the sender and delete all copies. >
