[ 
https://issues.apache.org/jira/browse/MAHOUT-559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12969067#action_12969067
 ] 

Lance Norskog commented on MAHOUT-559:
--------------------------------------

Thanks for the comments. This version is just a sanity check on the concept. 
(Apparently the answer is yes.) I'll absorb your comments. I have more changes 
anyway. For example, javadocs.
bq. I need to move the CompareRecommenders class to examples/ since it depends 
on examples code, and is itself really an example
Yup.
bq. Looks like it depends on the "Estimating" recommenders - are those yours? 
They're not in the patch so may I just remove those methods?
Oops. Those are personal examples. I'm building this all in Eclipse so my paths 
are different than Maven.
bq. For OrderBasedRecommenderEvaluator I'd want to log results rather than 
write to PrintStream
These are for analysis in Excel, KNime, R etc. So, CSV is the right format.
bq. I might want to fix stuff like this? ((Object)null).hashCode(); to throws 
IllegalStateException or something, etc.
Debugging code. Will remove.


> Compare Recommender output by order of recommendations.
> -------------------------------------------------------
>
>                 Key: MAHOUT-559
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-559
>             Project: Mahout
>          Issue Type: New Feature
>          Components: Collaborative Filtering
>            Reporter: Lance Norskog
>         Attachments: OrderBasedRecommenderEvaluator.patch
>
>
> The existing RecommenderEvaluator 
> (AverageAbsoluteDifferenceRecommenderEvaluator.java) has a very limited API. 
> It evaluates a Recommender's performance on a training v.s. test scenario. It 
> does not allow comparing the outputs of different recommenders against the 
> same data model. Also, I could not figure out how its comparison criteria.
> OrderBasedRecommenderEvaluator compares the output of two recommenders. It 
> only checks the order of the items in the recommendations, ignoring the 
> returned preference values. 

-- 
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