Space: Apache Mahout (https://cwiki.apache.org/confluence/display/MAHOUT)
Page: RecommendationExamples
(https://cwiki.apache.org/confluence/display/MAHOUT/RecommendationExamples)
Edited by yangy:
---------------------------------------------------------------------
h1. Introduction
This quick start page describes how to run the recommendation examples provided
by Mahout. Mahout comes with four recommendation mining examples. They are
based on netflixx, jester, grouplens and bookcrossing respectively.
h1. Steps
h2. Testing it on one single machine
In the examples directory type:
{code}
mvn -q exec:java
-Dexec.mainClass="org.apache.mahout.cf.taste.example.bookcrossing.BookCrossingRecommenderEvaluatorRunner"
-Dexec.args="<OPTIONS>"
mvn -q exec:java
-Dexec.mainClass="org.apache.mahout.cf.taste.example.netflix.NetflixRecommenderEvaluatorRunner"
-Dexec.args="<OPTIONS>"
mvn -q exec:java
-Dexec.mainClass="org.apache.mahout.cf.taste.example.netflix.TransposeToByUser"
-Dexec.args="<OPTIONS>"
mvn -q exec:java
-Dexec.mainClass="org.apache.mahout.cf.taste.example.jester.JesterRecommenderEvaluatorRunner"
-Dexec.args="<OPTIONS>"
mvn -q exec:java
-Dexec.mainClass="org.apache.mahout.cf.taste.example.grouplens.GroupLensRecommenderEvaluatorRunner"
-Dexec.args="<OPTIONS>"
{code}
Here, the command line options need only be:
{code}
-i [input file]
{code}
Note that the GroupLens example is designed for the "1 million" data set,
available at http://www.grouplens.org/node/73 . And the "input file" above is
the ratings.dat contained in the zipfile from the data set . This file has an
unusual format and so has a special parser. The example code here can be easily
modified to use a regular FileDataModel and thus work on more standard input,
including the other data sets available at this site.
Change your notification preferences:
https://cwiki.apache.org/confluence/users/viewnotifications.action
