Icon

Market Basket Analysis - Mine rules

<p><strong>Market Basket Analysis - Mine rules</strong></p><p>This workflow reads transaction and product data, then uses the Borgelt version of the Apriori algorithm to discover patterns of items frequently bought together.</p><p>The workflow allows you to set a minimum support threshold (how common a pattern must be) to control which rules are found. After generating the rules, it enriches them by attaching product names and price statistics to make the results more understandable and useful. Finally, the complete set of association rules is saved to a file for further analysis or reporting.</p>

Market Basket Analysis - Mine rules


This workflow reads transaction and product data, then uses the Borgelt version of the Apriori algorithm to discover patterns of items frequently bought together. The workflow allows you to set a minimum support threshold (how common a pattern must be) to control which rules are found. After generating the rules, it enriches them by attaching product names and price statistics to make the results more understandable and useful. Finally, the complete set of association rules is saved to a file for further analysis or reporting.

Data Access & Preparation

Historical basket data:

  1. Transactions

  2. Products

Compute Association Rules

Match Antecedent/Consequent Product ID with Product Name

Computer Revenue for Recommended Items

Save Association Rules

Calculatingrevenues followingrecommendation as price(consequent) * support(rule)
Expression
Associate consequentproduct ID to product name
Consequents to product name
Translate antecedentcollection to productname concatenations
Antecedents to product name
Calculate association rules("A priori" algorithm)
Association Rule Learner (Borgelt)
Transactions:1 row = 1 transactionproduct IDs bought for each transaction
Table Reader
Write association rulesto a file
Table Writer
Data preparation
Products:ID, price, productName
Table Reader
Double-click to set parameter for algorithm
Set parameter

Nodes

Extensions

Links