Can you please suggest me a good model-based learning algorithm to recommend items to the user? Is there any open source implementation available on model based learning algorithm? I am sure Apache Mahout doesn't implemented any model based learning algorithms.
-
1$\begingroup$ Mahout can do even SVMs and RFs, they're very model-based. $\endgroup$– user88Commented Jun 8, 2011 at 8:43
-
$\begingroup$ From Mahout document it says, Mahout supports both memory-based, item-based recommender systems, slope one recommenders, and a couple other experimental implementations. It does not currently support model-based recommenders. cwiki.apache.org/confluence/display/MAHOUT/… $\endgroup$– Ananth DuariCommented Jun 8, 2011 at 9:59
-
$\begingroup$ @Ananth So you must only manually make a recommender from a model. $\endgroup$– user88Commented Jun 8, 2011 at 10:45
-
$\begingroup$ well it is not with Mahout now. But is there are any other opensource been implemented on that? $\endgroup$– Ananth DuariCommented Jun 8, 2011 at 10:48
-
$\begingroup$ There is a reason why Mahout lacks them -- it is just too hard to consider all the possibilities and optimize such recommender. You won't find an app with "Make model recommender" button -- this is just a work for a human do design such system personalized to your case. $\endgroup$– user88Commented Jun 8, 2011 at 11:01
2 Answers
If you set your recommender system up as a matrix factorization, then there will be plenty of Mahout experience out there, e.g. https://mahout.apache.org/users/recommender/matrix-factorization.html or http://de.slideshare.net/datasciencelondon/matrix-factorization-mahout-special-session
If you have a problem with expressing a recommender system like this, then have a look at this question & answer: https://datascience.stackexchange.com/questions/10114/using-movielens-dataset-build-recommendation-engine/10118#10118
Ask please if you have further questions. I won't be able to help you with Mahout but with the MF part.
Pat
There are several opensource offerings including Mahout, Lenskit, Crab, Duine,etc.
Here is an overview from ibm.com: