Title: A market basket analysis of seven retail branches in Kyrgyzstan using an Apriori algorithm

Authors: Nematullah Wahidi; Rita Ismailova

Addresses: Computer Engineering Department, Kyrgyz-Turkish Manas University, 720038 Bishkek, Kyrgyzstan ' Computer Engineering Department, Kyrgyz-Turkish Manas University, 720038 Bishkek, Kyrgyzstan

Abstract: The proliferation of online trading platforms compelled businesses to delve into the analysis of customer behaviour. Therefore, this study seeks to scrutinise customer behaviour in Kyrgyzstan, a previously unexamined market, aiming to augment supplier revenue, service quality, and customer satisfaction. The transaction data employed in this investigation was sourced from the retail sector and processed across seven distinct regions. The analysis was done utilising the Apriori algorithm. Our findings reveal that Bishkek city, the premier sales region, yielded 866 association rules from a dataset of 308,998 records. These rules demonstrate robust connections among items, indicating an 83% probability of relationships between the consumption patterns of various products. The experimental verification highlights significant patterns and associations with a high degree of confidence. The discerned insights can profoundly inform targeted marketing initiatives, optimise inventory management practices, elevate customer experiences, and promise to foster growth within the dynamic landscape of the retail sector.

Keywords: association rule mining; ARM; incremental data mining; data mining; Apriori algorithm; Kyrgyzstan.

DOI: 10.1504/IJBIDM.2025.143939

International Journal of Business Intelligence and Data Mining, 2025 Vol.26 No.1/2, pp.236 - 255

Received: 26 Feb 2024
Accepted: 16 Sep 2024

Published online: 14 Jan 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article