Identifying Characteristics of Dietary Guidance Methods Through Qualitative Analysis
摘要
Continuous support for improving lifestyle habits is effective in preventing metabolic syndrome. However, providing sustained support by healthcare professionals imposes a significant burden on them. With the increasing use of large language models (LLMs) capable of efficiently generating messages, incorporating the characteristics of dietary guidance by healthcare professionals into LLM prompts may enable the generation of messages with the same motivational effect as professional guidance. To identify these key characteristics, we conducted a dietary-improvement experiment involving the LOCABO Challenge dietary-guidance program and collected data on dietary advice. Through qualitative analysis, we elucidated the process by which healthcare professionals review participants’ dietary records, formulate advice, and use motivational strategies in dietary guidance. We also explored potential approaches for incorporating our findings into the LLM design.