Article: Personalised cultural creative product design using user profile and personalised data diffusion model Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.49 pp.19 - 35 Abstract: Addressing the issue that current cultural and creative product generation methods fail to account for user emotional needs, resulting in poor image generation outcomes, this paper first employs natural language processing algorithms to automatically segment user profiles and extract demand characteristics. Deep learning algorithms are introduced to analyse the sentiment behind user demands, thereby identifying emotional inclinations expressed by users. Building upon this foundation, a novel residual block architecture is designed with a diffusion model as the core network. The noise estimation network is enhanced by incorporating a convolutional block attention module. By integrating conditional control and user profiles as the control network, the approach effectively generates cultural and creative product images that align with users' emotional expectations. Experimental results demonstrate that the proposed method achieves at least an 8.63% improvement in peak signal-to-noise ratio, enabling the generation of high-quality cultural and creative product images. Inderscience Publishers - linking academia, business and industry through research

Title: Personalised cultural creative product design using user profile and personalised data diffusion model

Authors: Dongxu Yang; Yongxin Guo; Ronghui Liu

Addresses: School of Art and Design, Henan University of Urban Construction, Pingdingshan, 467000, China ' School of Art and Design, Henan University of Urban Construction, Pingdingshan, 467000, China ' School of Computer and Data Science, Henan University of Urban Construction, Pingdingshan, 467000, China

Abstract: Addressing the issue that current cultural and creative product generation methods fail to account for user emotional needs, resulting in poor image generation outcomes, this paper first employs natural language processing algorithms to automatically segment user profiles and extract demand characteristics. Deep learning algorithms are introduced to analyse the sentiment behind user demands, thereby identifying emotional inclinations expressed by users. Building upon this foundation, a novel residual block architecture is designed with a diffusion model as the core network. The noise estimation network is enhanced by incorporating a convolutional block attention module. By integrating conditional control and user profiles as the control network, the approach effectively generates cultural and creative product images that align with users' emotional expectations. Experimental results demonstrate that the proposed method achieves at least an 8.63% improvement in peak signal-to-noise ratio, enabling the generation of high-quality cultural and creative product images.

Keywords: cultural and creative product generation; user profiling; conditional diffusion model; attention mechanism; natural language processing.

DOI: 10.1504/IJICT.2025.151058

International Journal of Information and Communication Technology, 2025 Vol.26 No.49, pp.19 - 35

Received: 22 Sep 2025
Accepted: 25 Oct 2025
Published online: 12 Jan 2026 *

Title: Personalised cultural creative product design using user profile and personalised data diffusion model

Keep up-to-date