A model for the intelligent analysis of customer purchasing behavior in E-COMMERCE systems

Вантажиться...
Ескіз

Назва журналу

Номер ISSN

Назва тому

Видавець

Анотація

This article is devoted to the development of a conceptual model of a decision support system for analyzing customer purchasing behavior in the field of e-commerce. It demonstrates the limitations of descriptive analytics and traditional reporting, which record only aggregate performance metrics, and proves the need to implement data mining and machine learning methods. The authors propose a comprehensive DSS that integrates RFM analysis for the initial assessment of customer value, the K-means algorithm for behavioral segmentation in a multidimensional feature space, and the Random Forest ensemble algorithm for predicting the risk of customer churn. An extended mathematical formulation of the problem is proposed, which formalizes the relationship between input transaction data, clustering procedures, the estimation of churn probability, and the automatic generation of targeted management interventions. The article also highlights key methodological limitations and risks of the model, including the “cold start” problem for new users, invariance to exogenous market factors, and the phenomenon of conceptual drift. The practical application of the model facilitates the transition from retrospective data to predictive loyalty management and enhances the effectiveness of marketing activities.

Опис

Бібліографічний опис

Brynza N. O. A model for the intelligent analysis of customer purchasing behavior in E-COMMERCE systems / N. O. Brynza, D. D. Koliesnik // SWorldJournal. – 2026. – Iss. 39. – Part 1. – P. 3-18.

Зібрання

Endorsement

Review

Supplemented By

Referenced By