Greetings all,

I recently got a patch accepted into Mahout (MAHOUT-202) that includes
an implementation of their DataModel for HBase (aptly named
HBaseDataModel). The DataModel specifies a means to store user ratings
(preferences) for items. These ratings are then used for computing
recommendations. Search for "collaborative filtering" if you want to
learn more (it's really cool stuff).

Anyways, I'm sending this out to you all to get some HBase-savvy
feedback. Here is the JIRA issue
https://issues.apache.org/jira/browse/MAHOUT-202

Cheers!
--
David Arthur

On Wed, Oct 3, 2012 at 5:49 PM, David Arthur <[EMAIL PROTECTED]> wrote:
> I recently got a patch accepted into Mahout (MAHOUT-202) that includes
> an implementation of their DataModel for HBase (aptly named
> HBaseDataModel). The DataModel specifies a means to store user ratings
> (preferences) for items. These ratings are then used for computing
> recommendations. Search for "collaborative filtering" if you want to
> learn more (it's really cool stuff).
>
> Anyways, I'm sending this out to you all to get some HBase-savvy
> feedback. Here is the JIRA issue
> https://issues.apache.org/jira/browse/MAHOUT-202
>

I read through the issue.  Seems like you did your homework David.
Good on you.  Thanks for adding it.   Hopefully it'll get used.
Running HBase standalone, I think you will might need more than just
hbase.jar and protobuf.  There'd be zookeeper, etc., stuff I' not
think already in Mahout.  Good on you David.

St.Ack