Using Association Rules for Market Basket Analysis

Using Association Rules for Market Basket Analysis

Data Reading

3 files:

• Association rules

• Products info

• Read current basket data

The files are located in TheData/Baskets

For the basket data, in this case we used a Table Creator node. You can read your current basket data from your database or from whatever source.

Double-click the metanode to see the subworkflow.

Data Processing

• Find the best match for the current basket using the Subset Matcher node

• Find top matching rules in terms of highest support, highest confidence, and highest revenue

• Calculate total price of current basket

• Append price of current basket

• Split the two top recommendations in two different tables

Double-click the metanode to see the subworkflow.

Visualize

Display the two recommendations in an interactive table view.

Task

Apply the pre-built association rules to a new set of data