Title: Use of machine learning for classifying manufacturing companies based on their digital transformation levels

Authors: Ece Acar; Gorkem Sariyer

Addresses: Department of Business Administration, Faculty of Business, Yasar University, Izmir, Turkey ' Department of Business Administration, Faculty of Business, Yasar University, Izmir, Turkey

Abstract: The transformative role of machine learning technology in promoting technological innovation leading sustainable growth is becoming increasingly significant in today's business era. In this study, we implemented machine learning technology to classify the companies according to their digital transformation levels. We used manufacturing companies in Borsa Istanbul (BIST) index as the sample. We constructed a digital transformation level index based on text analysis to measure the frequency of keywords related to digital transformation. We used the sampled companies' financial, sustainability, corporate governance performance and research & development (R&D) expenditures to model their digitalisation levels. We observed that between the various machine learning algorithms, with 82% accuracy, Random Forest outperformed the others. We also showed that while R&D expenditure was the most important feature, financial performance-related features were also significant. Thus, we concluded that companies with higher financial performances, especially those making more expenditures for R&D activities, have higher digital transformation levels.

Keywords: digital transformation; financial performance; R&D expenditure; machine learning; classification.

DOI: 10.1504/IJIE.2025.147695

International Journal of Intelligent Enterprise, 2025 Vol.12 No.3/4, pp.305 - 320

Received: 18 Dec 2023
Accepted: 30 Jul 2024
Published online: 25 Jul 2025 *

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