mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sean Owen <sro...@gmail.com>
Subject Re: Questions about PearsonCorrelation on a example
Date Tue, 23 Jun 2009 23:08:34 GMT
If you are interested in recommendations based on a clustering
approach, look at TreeClusteringRecommender. I never really 'finished'
it. It works but is slow. At the least it may be of interest, as a
starting point for you, since it does produce clusters as part of the
computation, which it sounds like you want for other reasons.

Sean

On Tue, Jun 23, 2009 at 6:52 PM, charlysf<charles.ruelle@gmail.com> wrote:
>
> Thanks.
> This recommendation is my explicit basis recommendation, for the implicit, I
> don't have this problem, so I will start to recommend based on the
> implicit... and store more explicit data and after to merge the both.
>
> Now, I have a similarity based on implicit & explicit ratings, and I would
> like to make some clusters of users, which algorithm do you recommand to me
> ?
> I would like to do the same after for items, and to have some
> recommendations based on clusters.

Mime
View raw message