Analysis of Customer Expense using Market Basket Analysis Model with Hash-Based Algorithm and Association Rules

Abstract

The Market Basket Analysis is a method to analyze consumer spending patterns. The method use data that is retrieved from the transaction database, then processed to obtain information. Normally, the data transaction is obtained from supermarket or any retail business. The transaction database does not serve management any information because it only contains raw data. Therefore, market basket analysis is conducted to analyze consumer's patterns. In this study, the algorithm used is a hash-based algorithm. The algorithm reduces the number of candidate item-sets at the beginning. The results of this study found 2 frequent items that is being bought together, namely bread and butter with support value of 96. Then eggs, bread and butter with support value of 97. Based on the results of the implementation of hash-based algorithm and association rules, we can conclude that managers or any stakeholder is suggested to use promotion on the products above. Those items that are being frequently bought together need to be placed in the same shelf or near each other to accommodate consumer for convenient purposes.