Mining User Requirement Scenarios and Generating Design Solutions for Rehabilitation Aids Based on Large Language Models
Xinyu Pan, Jie Kai Gong, Sijie Wen, Weibin Zhuang, Xinyu Li · 2024
The development of the rehabilitation aids industry for the disabled has been pivotal in recent years, particularly in the personalized design of lower limb rehabilitation aids. Facing challenges in meeting individualized demands in design practice and the information gap between medical professionals and users, we propose a design Knowledge Graph (KG) method based on the Function-Behavior-Structure (FBS) model. This approach utilizes open-source large language models (LLMs) and fine-tunes them with instruction data generated by self-instructions to improve the accuracy of user requirements mining. The method aims to enhance the personalization and innovation of rehabilitation aids design through the integration of KG and LLM, effectively narrowing the cognitive gap between service providers and users. The anticipated results of the study are expected to promote efficient innovation in rehabilitation aids design, better meeting the needs of the disabled community.