Yes, getCandidateItems() needs to be modified as well.
--sebastian
On 05/02/2014 10:39 AM, dodi hakim wrote:
Ok, but these methods are called in many places. Furthermore, the
recommend() is only the entry point to call getCandidateItem() method
in AbstractCandidateItemsStrategy class, so we still need to add a
parameter in the getCandidateItem() method.
On 2 May 2014 17:45, Sebastian Schelter <[email protected]> wrote:
I think it would be better to add a parameter to the recommend() methods
so that this can be decided on a per invocation basis.
--sebastian
On 05/02/2014 09:40 AM, dodi hakim wrote:
Thanks,
My proposed solution for this is to add a method in Recommender interface
(e.g isUserItemInclude) and pass a Recommender class in method
getCandidateItem in both CandidateItemStrategy and
MostSimilarItemsCandidateItemsStrategy interfaces. For example, in
method
doGetCandidateItems of PreferredItemsNeighborhoodCandidateItemsStrategy
class - that implements these two interfaces, it has line
possibleItemsIDs.removeAll(preferredItemIDs) to remove user's preferable
items.
In addition, we need to list which Recommender classes that including/not
user's item in returned recommendation.
Any advice and suggestions will be greatly appreciate.
Thanks,
Dodi
On 2 May 2014 16:53, Sebastian Schelter (JIRA) <[email protected]> wrote:
[
https://issues.apache.org/jira/browse/MAHOUT-1428?page=
com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel]
Sebastian Schelter reopened MAHOUT-1428:
----------------------------------------
here you are
Recommending already consumed items
-----------------------------------
Key: MAHOUT-1428
URL: https://issues.apache.org/jira/browse/MAHOUT-1428
Project: Mahout
Issue Type: Bug
Components: Collaborative Filtering
Reporter: Mario Levitin
Fix For: 1.0
Mahout does not recommend items which are already consumed by the user.
For example,
In the getAllOtherItems method of GenericUserBasedRecommender class
there is the following line
possibleItemIDs.removeAll(dataModel.getItemIDsFromUser(theUserID));
which removes user's items from the possibleItemIDs to prevent these
items from being recommended to the user. This is ok for many
recommendation cases but for many other cases it is not.
The Recommender classes (I mean all of them, NN-based and SVD-based as
well as hadoop and non-hadoop versions) might have a parameter for this
for
excluding or not excluding user items in the returned recommendations.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
