Comparative Analysis on Association Rule Mining (ARM) Algorithms Using Market Basket Analysis Approach

Abstract

Efficient monitoring and controlling of products in market cycle is the break-even in business venture such as grocery shops. Data of previous sales activities in grocery stores are majorly used for predicting customer purchasing power and behavioral change. This research will perform a comparative analysis review on Association rule mining models (ECLAT, Apriori and FP Growth) to determine the algorithm best fit for application implementation. This classifies the items that occur together frequently in customer transactions. A web-based application is developed and modelled to implement the best algorithm in the software application. This will be achieved using system object-oriented design analysis and the development tools via python flask framework for web backend and interface. Its benefit is deciding on market items, customers to appreciate, item categorization, and importantly uncovering the relationship, correlation, on items within a market and frequent item set in general grocery store purchased by customers.