You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mahout.apache.org by "Sean Owen (JIRA)" <ji...@apache.org> on 2010/12/05 13:27:11 UTC
[jira] Commented: (MAHOUT-558) Extend ItembasedRecommender to offer
different "exclusion modes" when computing most similar items to a
collection of input items
[ https://issues.apache.org/jira/browse/MAHOUT-558?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12966931#action_12966931 ]
Sean Owen commented on MAHOUT-558:
----------------------------------
One initial comment is, shouldn't this change affect not just MultiMostSimilarEstimator, but MostSimilarEstimator? I think it would have an analogous change in the user-based recommender too for consistency.
Let me present a small example to make sure I understand. The estimated preference for some new thing X is based on the user's existing preferences for A, B and C. Right now, if X has no defined similarity (not 0, but uncomputable or undefined) to A, B or C, then X cannot be recommended -- there's simply no estimate.
Right now, if X has a defined simiarity to *any* of A, B or C then some estimate can be made. The undefined relations are ignored. This is, I think, a decent intepretation of the stock item-based recommender approach; it does "unfairly" favor obscurer items.
If I read this change right, it would need for X to have a resemblance to *all* of A, B and C to produce any estimate. That sounds like quite a strict rule in practice -- would it not fail to produce recommendations in many cases? Is that the intent and if so what's the real-world use case for this.
> Extend ItembasedRecommender to offer different "exclusion modes" when computing most similar items to a collection of input items
> ---------------------------------------------------------------------------------------------------------------------------------
>
> Key: MAHOUT-558
> URL: https://issues.apache.org/jira/browse/MAHOUT-558
> Project: Mahout
> Issue Type: New Feature
> Components: Collaborative Filtering
> Affects Versions: 0.5
> Reporter: Sebastian Schelter
> Attachments: MAHOUT-558.patch
>
>
> GenericItembased Recommender currently excludes all items that are not similar at least one of the input items when computing the most similar items to a collection of items. We should introduce a way to have the user decide whether he/she wants this behavior or he/she wants to have all items included that are similar to at least one of the input items, which is more useful in practice in my experience.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.