Are there any recommender algorithms designed for micro-sharding the
data model? The use case would be a mobile app that stores only a data
model for the phone owner.
It seems like a user-user recommender does not need data for all
users; nearby users plus some background noise should be enough to
Yes. That is quite reasonably possible. It isn't really micro-sharding
since it will be different for every user rather than being a universal
sharding of all users.
On Tue, Aug 16, 2011 at 8:35 PM, Lance Norskog wrote:
> Are there any recommender algorithms designed for micro-sharding the
> d
Consider the recommender as two separate datasets:
1) a "useful number" of similar users as a neighborhood, and
2) a lot of more distant weights summarized and factored in somehow.
The new graph triangle-finder code would find all neighborhoods for
all users in one job.
On Tue, Aug 16, 2011 at 10
Yes... though this is almost identical to just retrieving recommendations
from an external server in the first place!
If you download the user's neighborhood, including things those users like,
you have effectively downloaded a list of all recommendable items, and info
to rank them.
This isn't en
Much larger item-item matrices should be possible with compressed
representations. After all, mobile devices are pretty big any more.
On Wed, Aug 17, 2011 at 1:36 AM, Sean Owen wrote:
> I do know of one company that loaded an entire, tiny recommender onto their
> mobile app. It was item-based a
