Article: Generative adversarial networks for simulating emotional resonance in industrial product design Journal: International Journal of Simulation and Process Modelling (IJSPM) 2026 Vol.23 No.2 pp.90 - 102 Abstract: This paper addresses the lack of emotion-oriented simulation in industrial product design by proposing a novel generative adversarial network framework integrated with a quantifiable emotional model. The core of this approach is an emotion-attention mechanism that dynamically guides the form evolution process. Emotional features are first extracted from e-commerce reviews and modelled via an improved support vector regression algorithm to establish a quantifiable mapping between design elements and user emotions. This emotional model is then integrated into a GAN through a multi-head component attention module, which simulates product form evolution by explicitly weighting the contribution of each component to the target emotional resonance. Experimental results demonstrate the effectiveness of this simulation, with the Fréchet inception distance reduced by at least 38.06%, enabling the generation of industrial products that accurately align with user emotional needs. Inderscience Publishers - linking academia, business and industry through research

Title: Generative adversarial networks for simulating emotional resonance in industrial product design

Authors: Jie Hu

Addresses: College of Design and Art, Jingdezhen Ceramic University, Jingdezhen, 333001, China

Abstract: This paper addresses the lack of emotion-oriented simulation in industrial product design by proposing a novel generative adversarial network framework integrated with a quantifiable emotional model. The core of this approach is an emotion-attention mechanism that dynamically guides the form evolution process. Emotional features are first extracted from e-commerce reviews and modelled via an improved support vector regression algorithm to establish a quantifiable mapping between design elements and user emotions. This emotional model is then integrated into a GAN through a multi-head component attention module, which simulates product form evolution by explicitly weighting the contribution of each component to the target emotional resonance. Experimental results demonstrate the effectiveness of this simulation, with the Fréchet inception distance reduced by at least 38.06%, enabling the generation of industrial products that accurately align with user emotional needs.

Keywords: product form simulation; process modelling; emotional resonance; generative adversarial networks; GAN; support vector regression; SVR.

DOI: 10.1504/IJSPM.2026.154273

International Journal of Simulation and Process Modelling, 2026 Vol.23 No.2, pp.90 - 102

Received: 08 Dec 2025
Accepted: 12 Mar 2026
Published online: 18 Jun 2026 *

Title: Generative adversarial networks for simulating emotional resonance in industrial product design

Keep up-to-date