Monday, December 19, 2016

Advanced Analytics: Association Rule Mining on OracleDV using Custom R-Scripts

Association Rule Mining is a common technique used to find associations between many variables. It is intended to identify strong rules existing in data using some measures of interestingness. It is often used by grocery stores to perform Market Basket Analysis(MBA), and used by online stores to provide suggestions for purchases. 

Do you have a transactional detaset with you and would like to perform Association Rule Mining on it? You can do it very easily on OracleDV using Rule mining Custom R-Script. The R-script returns association rules in a tabular format with Support, Confidence and Lift associated with each rule. This list of association rules can also be exported/downloaded to excel format. You can also use the script in Dataflows by saving the Rule set generated from your data and consume it as a source in the dataflow. In this blog we will discuss how you can deploy this R-Script and perform Rule Mining on OracleDV Desktop.

Steps to deploy:

1) Install Advanced Analytics feature in Oracle DV by clicking on the below icon. This will install Oracle R deployment. Alternatively you can install Advanced Analytics by running install_advanced_analytics.cmd present in <DV_INSTALL_DIRECTORY>

2) If not installed arules R-Package already, please install it using following instructions
    Open R console(double click Rgui.exe present in <Advanced_Analytics_Install_Dir>\bin\x64) and
    install arules Package.
    Following are the R commands to install:
     Set Proxy:
        $ Sys.setenv(http_proxy="<your_proxy_host>:<port_number>")
           set proxy appropriate to your network config.
     Install Package:
        $ install.packages("arules")

3) Download from OracleBI Public Store and unzip it.
4) Copy R.RuleMining.xml to <DV_INSTALL_DIRECTORY>\OracleBI1\bifoundation\advanced_analytics\script_repository
5) Import the .dva project to Oracle DV. Password for the .dva is Admin123

Here is a snapshot:

No comments:

Post a Comment