Read the full-text of this article for free

Title: Marketing credit evaluation model of B2B e-commerce enterprises based on improved CNN algorithm

Authors: Jianhao He; Fangliang Zhang

Addresses: School of Supply Chain Management, Ningbo Polytechnic University, Ningbo, 315800, China ' Nottingham University Business School China, University of Nottingham, Ningbo, China

Abstract: The traditional credit evaluation model has the disadvantages of difficult data mining and low evaluation accuracy. This study constructs an improved convolutional neural network and applies it to the marketing credit assessment technique used by e-commerce businesses. The results showed that the area under the accuracy-recall curve of this study, random forest, and decision tree were 0.83, 0.74, and 0.65, respectively. The area of the convolutional neural network under the curve of number of iterations was 0.65. The convolutional neural network completed the iteration when the iteration number was 198 times. The random forest model completed the iteration when the iteration number was 239 times. The decision tree model completed the iteration when the iteration number was 594. Thus, the suggested method owns better accuracy and robustness.

Keywords: neural network; marketing; credit evaluation; convolutional neural network; CNN; model.

DOI: 10.1504/IJICT.2026.151488

International Journal of Information and Communication Technology, 2026 Vol.27 No.2, pp.1 - 18

Received: 31 Jul 2025
Accepted: 03 Nov 2025
Published online: 02 Feb 2026 *