Generative adversarial network model-based knowledge recommendation with knowledge graph in product design
Zhenyong Wu, Jianxin Tan, Hanzhao Wu, Wendi Zha · Journal of Engineering Design · 2025
As the integration of manufacturing with advanced information technology progresses, data and knowledge have become foundational pillars, ensuring a seamless transition to intelligent manufacturing. Throughout the whole lifecycle, products generate a vast amount of industrial data and knowledge. In innovative design, leveraging previous engineering empirical knowledge can play a pivotal role. The paper proposes a three-layered framework leveraging knowledge graph methodologies for product design, aiming to enhance knowledge reuse efficiency and address knowledge discovery challenges, a knowledge graph-based integration method is presented, integrating with semantic automatic annotation and keyword extraction. A GAN-based knowledge recommendation model is developed to integrate both user context information and problem context information into collaborative training. To validate the proposed approach, a specific case study validates the framework's practicality and effectiveness.