Market Basket Analysis on TaFeng Grocery Store

This is my first mini-project for our Big Data and Cloud Computing class.

Summary

One of the common challenges for companies is how to optimize their inventory. Although a number of inventory management techniques and methods already exist, these do not incorporate the interactions/relationship between products. This study aims to implement association pattern mining to discover relationships between pairs of products purchased together that will further enhance inventory management. A publicly available grocery dataset from TaFeng grocery that contains 817,741 transactions from November 2000 to February 2001 was consolidated, cleaned, pre-processed, examined, and analyzed using frequent itemset mining and association pattern mining. 560 frequent itemsets are mined from the shopping transactions using the ECLAT algorithm, and 22 association rules are generated. The results uncovered related products that are frequently bought together by customers. From the results identified, frequent itemsets must never be out of stock else the company will lose the opportunity to maximize profit.