Market Basket Analysis on Telecommunications Prepaid Packages using the Apriori Algorithm

Data package purchase transactions with telecommunications operators are one of the company’s revenue generators. Purchasing operator packages can be done either through applications, e-commerce or distributors and outlets, where customers get a lot of exposure to telecommunications package offers. Telecommunications operators have the problem of how to be able to provide more appropriate offers to customers by determining which products are suitable to be offered on their channels. The purpose of this research is to analyze the purchase transaction data of operator data packages and use data mining to find patterns of package purchases that can be used to increase purchase transactions. This research is limited to processing transaction data and modeling with the results of the modeling algorithm which will be limited to product recommendations that are suitable for customers by taking the product package variations offered. The data mining method used uses Market Basket Analysis with the Apriori algorithm which produces predictions of package combinations that will be purchased by customers by utilizing a sample customer purchase transaction history data in March 2023. The model produces 19 product combinations to be offered to customers with the main product being a data package small quota of 500 MegaBytes combined with telephone package. The Main Combination that is formed shows customer behavior consisting of: a combination of small quota data packages with large quota telephone packages, combination of large quota data packages with small quota telephone packages, and digital packages combined with small quota data packages. Evaluation is carried out using support parameter limits >= 1%, confidence value >= 1% and lift ratio >= 1 of the achievement of the algorithm.