Abstract

Background: The rapid advancement of Artificial Intelligence (AI) presents new opportunities for protecting cultural heritage and enhancing of industrial growth. As a result, the construction of traditional toy genetic maps and AI-driven design innovation has become essential for the ongoing cultural and industrial development. Purpose: This study aims to: 1) develop a multi-dimensional cultural gene mapping model for traditional toys that covers morphological, craft, and symbolic dimensions; 2) propose an AI-driven framework to translate tacit cultural knowledge into explicit digital expressions; 3) clarify the transformative role of AI in reconstructing the traditional toy industry ecosystem. Methods: This research adopts a mixed-method approach based on cultural gene theory and knowledge engineering. It analyzes over 120 typical motifs and 30 craftsmanship techniques. By employing machine learning algorithms such as Convolutional Neural Networks (CNN) for form recognition and Long Short-Term Memory (LSTM) networks for craft process modeling, the study aims to build digital gene libraries. The findings are validated through case studies of toy categories including Niren Zhang clay figures, fieldwork conducted at eight heritage sites, and interviews with 20 industry experts. Results: The study develops a three-layer gene model consisting of morphological genes, craft genes, and symbolic genes. The AI framework enhances design iteration efficiency by 42% and improves cultural symbol accuracy by 35%. It also fosters democratized design with 62% of innovative patterns emerging from non-experts in pilot projects. Conclusion: This research establishes a framework in which AI serves as a medium for cultural re-creation rather than merely a tool. The cultural gene mapping model offers a scalable solution for digitizing intangible heritage, while the design framework bridges traditional craftsmanship with modern innovation.