Chapter 8

Title: Basic Association Rule Mining in RapidMiner

Summary

Chapters 8 to 10 describe three different approaches to building recommender systems. Product recommendations in online shops like Amazon increase the sales volume per customer by cross-selling, i.e., by selling more products per customer by recommending products that the customer may also like and buy. The recommendations can be based on product combinations frequently observed in market baskets in the past. Products that co-occurred in many purchases in the past are assumed to be also bought together frequently in the future. Chapter 8 describes how to generate such association rules for product recommendations from shopping cart data using the FP-Growth algorithm. Along the way, this chapter also explains how to import product sales data from CSV files and from retailers’ databases and how to handle data quality issues and missing values.

Dataset & Processes: Click here to download