The research involves the application of the "Diabetes Smart Helper" LINE@ Diabetes Health Education Chatbot for intervention measures. This application includes AI automatic message responses and provides push notifications of diabetes-related medical and healthcare knowledge to ensure that the public receives accurate medical information (Chinese Diabetes Health Education Association, 2021).

Upon opening the LINE application, the research subjects need to enter "Diabetes Smart Helper" in the LINE@ Diabetes Health Education Chatbot. Once added to the program, the LINE interface will directly display six small windows for diabetes symptoms/complications, diet, medication, exercise, mental health, and others, facilitating self-selection by users.

This study employs the Chinese Version Diabetes Management Self-Efficacy Scale (C-DMSES) developed by S. F. Wu et al. (2008) for assessment. The scale covers nutrition (9 items), exercise and weight control (4 items), medication management (3 items), and blood glucose and foot care (4 items). Scoring is based on an 11-point Likert scale, where 10 points indicate "very confident" and 0 points indicate "not confident at all." The total score ranges from 0 to 200, with higher scores indicating stronger self-efficacy, meaning that patients have greater confidence in performing diabetes self-care behaviors such as blood glucose monitoring, dietary management, control, and medication adherence. This scale is used for patients with type 2 diabetes in Taiwan, and the internal consistency reliability of the four subscales is as follows: dietary self-efficacy (α = .93), adherence to medical advice self-efficacy (α = .79), exercise and weight control self-efficacy (α = .81), and blood glucose mo

The Summary of Diabetes Self-Care Activities (SDSCA) published by Toobert et al. (2000) was used, which includes 25 questions covering six aspects, including diet, exercise, self-monitoring of blood glucose, smoking, self-care behavior counseling, and medication. The questionnaire was modified by Qiu (2002) to measure the self-care behavior of diabetes in Chinese. There are 10 structured questions, including foot care (3 questions), diet (2 questions), exercise (2 questions), medication (2 questions) and blood sugar monitoring (1 question). This scale monitors the daily self-care of diabetes in the past 7 days. The second question is a reverse question, and the other questions are scored in a positive way. The 7-day frequency method is used for calculation. The questionnaire scale is calculated from 0 to 7 points. Each question has a maximum score of 7 points and a minimum score of 0. For example, if the questions in the questionnaire are completed for 1 day, it represents 1 point, an

This study is based on the 2022 Clinical Care Guidelines for Type 2 Diabetes. Blood sugar control indicators include blood sugar level before meals, blood sugar level 2 hours after meals, and HbA1c value. It is recommended that good blood sugar control indicators for patients with type 2 diabetes are blood sugar level before meals maintained at 80-130 mg/dl, blood sugar level 2 hours after meals maintained at 80-160 mg/dl, and HbA1c value < 7.0% (Diabetes Society of the Republic of China, 2022). The above test values are from the medical records of the subjects.

The testing equipment used by the hospital in this case has passed ISO 1589:2012, with the certification number 1575, and meets the relevant standards. According to the requirements of the National Health Insurance Diabetes Medical Benefit Improvement Program, diabetic patients must undergo a blood test every 3 months. Patients must complete the blood test 3 to 7 days before outpatient consultation. The blood draw time of th