[ 
https://issues.apache.org/jira/browse/MAHOUT-389?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Sean Owen updated MAHOUT-389:
-----------------------------

    Issue Type: Improvement  (was: Bug)
      Priority: Minor  (was: Major)

I should add you could emulate the behavior you may desire in the 
non-distributed implementation with a combination of the 
UncenteredCosineSimilarity, and a PreferenceInferrer which always infers 0.

I don't see a simple way to make the distributed version emulate the 
non-distributed version.

While I think the non-distributed version has preferable semantics, neither is 
really "wrong". I suggesting leaving this as more of an improvement idea to 
figure out a way to unify them, by modifying the distributed version.

> UncenteredCosineSimilarity 
> ---------------------------
>
>                 Key: MAHOUT-389
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-389
>             Project: Mahout
>          Issue Type: Improvement
>          Components: Collaborative Filtering
>            Reporter: Sebastian Schelter
>            Priority: Minor
>         Attachments: MAHOUT-389.patch
>
>
> org.apache.mahout.cf.taste.impl.similarity.UncenteredCosineSimilarity only 
> computes the cosine distance between those components of the vectors where 
> both vectors have a value greater zero.
> This is inconsistent with the definition of the cosine (correct me if I'm 
> wrong) and is inconsistent with the distributed cosine similarity computation.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to