A Critical Review of Market Basket Analysis on Retail Dataset using Data Mining Techniques

Abstract

The associations between different products can be interpreted using a data mining technique called Market Basket Analysis (MBA). It contributes a vital role to determine the placement of goods, and the design of business strategies for retailers to attract consumer attraction and hence improve the businesses. The relationship among items can be deduced using association rules (AR) from retail datasets. As customer demand changes rapidly and thereby increasing transactional data. So, there is a need to use deep learning (DL) methods along association rules. Therefore, a review is conducted on MBA using AR and DL methods to mine and predict customer purchase patterns from large retail data sets. The objective of the paper is to assist researchers in the implementation of AR and DL methods while conducting MBA to overcome the challenges of large and frequently changing transactional data.