Spark xCLiMF

Spark implementation of the Extended Collaborative Less-isMore Filtering based in the python xCLiMF implementation. xCLiMF is a matrix factorization algorithm to optimize recommendation ranked lists.

Using

Example: Main.scala

Installation

Using gradle:

compile 'com.timotta:xclimf-spark:0.0.5'

Using maven:

<dependency>
  <groupId>com.timotta</groupId>
  <artifactId>xclimf-spark</artifactId>
  <version>0.0.5</version>
</dependency>

Requirements

• Java 8
• Scala 2.11.*
• Spark 2.1.1

Contribute

To test

gradle clean test

To run main with spark

gradle clean jar
$SPARK_HOME/bin/spark-submit --master local --class com.timotta.rec.xclimf.Main build/libs/xclimf-spark.jar

To upload new package

gradle -Prelease uploadArchives

To release

gradle -Prelease closeAndReleaseRepository