[
https://issues.apache.org/jira/browse/MAHOUT-1420?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13909997#comment-13909997
]
Pat Ferrel commented on MAHOUT-1420:
------------------------------------
The version I just downloaded doesn't have a ratings.dat but does have several
files with the right data. The version of the netflix data I have also will
work. One thing that has to be worked around it that the Solr recommender
doesn't work with preference weights, only boolean. So a simple r > 2.5 = 1; r
<=2.5 = 0 transform needs to be added.
Both look like they are from normalised DBs though so they don't look much like
logfiles. They will still work but will not really demonstrate what is
different about this pipeline. When using it, the items and users can have
human readable ids going in and coming out. The point isn't so much about human
readable as it is about using external ids in whatever form the user wants--not
Mahout integers.
hu...@email.com,iPad,purchase,time=awhile-ago,electronics,tablets,store.apple.com...
Or whatever.
> Add solr-recommender to examples
> --------------------------------
>
> Key: MAHOUT-1420
> URL: https://issues.apache.org/jira/browse/MAHOUT-1420
> Project: Mahout
> Issue Type: New Feature
> Components: Examples
> Affects Versions: 0.9
> Reporter: Andrew Musselman
> Assignee: Andrew Musselman
> Priority: Minor
> Fix For: 1.0
>
>
> Write a new example that builds a solr-recommender based on Pat's code at
> https://github.com/pferrel/solr-recommender and which has the glue scripts
> needed to pipe all the way from start(raw data) to finish(running web service
> and UI page).
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
