On Tue, Mar 2, 2010 at 10:42 AM, Claudio Martella
<[email protected]> wrote:
> Sean Owen wrote:
>> Yes, you're trying to solve a slightly simpler problem than
>> recommending items to users. You just want to find users similar to a
>> given user.
>>
>> Yes you di want the method UserBasedRecommender.mostSimilarUserIDs().
>> It's implemented at the moment by GenericUserBasedRecommender and
>> others
>>
>
> Wouldn't it be just easier and faster to directly use
> UserNeighborhood.getUserNeighborhood(userdID)?

You can do that too, yes. They're doing much the same thing, with
slight variation. For example the neighborhood calculation can be
merely sampled from among all users, and the one in the recommender
can use a Rescorer. They have slightly different needs.

Reply via email to