Association Rule Based Product Recommendations for Online Retail Customers

Below are the basket information of 3 different users. Make product recommendations for these baskets using the association rule. The product recommendations can be one or more. Derive the decision rules based on the 2010-2011 Germany customers.

• User 1's basket contains product ID: 21987
• User 2's basket contains product ID: 23235
• User 3's basket contains product ID: 22747

Dataset Story

The dataset named 'Online Retail II' includes online sales transactions of a UK-based retail company between 01/12/2009 and 09/12/2011. The product catalog of the company includes gift items, and it's known that most of its customers are wholesalers.

Variables:

• InvoiceNo: Invoice number (If this code starts with 'C', it indicates that the transaction was canceled)
• StockCode: Product code (Unique for each product)
• Description: Product name
• Quantity: Quantity of the product sold in each transaction
• InvoiceDate: Invoice date
• UnitPrice: Unit price of the product in the invoice (in GBP - British Pounds)
• CustomerID: Unique customer number
• Country: Country name