[jira] [Resolved] (MAHOUT-1464) Cooccurrence Analysis on Spark

["Pat Ferrel (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/MAHOUT-1464?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Pat Ferrel resolved MAHOUT-1464.
--------------------------------
       Resolution: Implemented
    Fix Version/s:     (was: 1.0)
                   0.10.0

SimilarityAnalysis has cooccurrence for columns, cross-cooccurrence too and row similarity with LLR filtering. No other "similarity" measure is implemented. Something like Cosine will require a different approach than A'A of AA' since it requires vector dot products while LLR only looks at occurrence counts--maybe something like the legacy mapreduce code. Deferring other similarity measures.

> Cooccurrence Analysis on Spark
> ------------------------------
>
>                 Key: MAHOUT-1464
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-1464
>             Project: Mahout
>          Issue Type: Improvement
>          Components: Collaborative Filtering
>         Environment: hadoop, spark
>            Reporter: Pat Ferrel
>            Assignee: Pat Ferrel
>              Labels: DSL, scala, spark
>             Fix For: 0.10.0
>
>         Attachments: MAHOUT-1464.patch, MAHOUT-1464.patch, MAHOUT-1464.patch, MAHOUT-1464.patch, MAHOUT-1464.patch, MAHOUT-1464.patch, run-spark-xrsj.sh
>
>
> Create a version of Cooccurrence Analysis (RowSimilarityJob with LLR) that runs on Spark. This should be compatible with Mahout Spark DRM DSL so a DRM can be used as input. 
> Ideally this would extend to cover MAHOUT-1422. This cross-cooccurrence has several applications including cross-action recommendations. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)