Abstract
The purpose of this study is to develop a sophisticated web-based artificial intelligence chatbot system designed to provide personalized dietary service for diabetic patients. According to a 2022 study, the prevalence of diabetes among individuals over 30 years old was 15.6% in 2020, identifying it as a significant societal issue with an increasing patient population. This study uses generative AI algorithms to tailor dietary recommendations for the elderly and various social classes, contributing to the maintenance of healthy eating habits and disease prevention. Through meticulous fine-tuning, the learning loss of the AI model was significantly reduced, nearing zero, demonstrating the chatbot's potential to offer precise dietary suggestions based on calorie intake and seasonal variations. As this technology adapts to diverse health conditions, ongoing research is crucial to enhance the accessibility of dietary information for the elderly, thereby promoting healthy eating practices and supporting disease prevention.