obtain a first question to a doctor, the first question being related to medical treatment of a patient and being entered by the patient via a graphical user interface (GUI) displayed on a terminal of the patient,

automatically without human intervention, generate at least one question candidate by inputting at least one keyword extracted from the first question entered via the GUI into a first machine-learning model, which outputs the at least one question candidate, wherein the first machine-learning model is trained using a first training set including input keywords and corresponding questions serving as first training data,

automatically without human intervention, convert the generated at least one question candidate output by the first machine-learning model into a second question by inputting the generated at least one question candidate into a second machine-learning model, which outputs the second question, the second question having equivalent content, but using a different expression, wherein the second machine-learning model is trained using a second training set including inappropriate questions as input data and corresponding appropriate questions as second training data;

transmit the second question output by the second machine-learning model over a network to the terminal of the patient for display on the terminal of the patient, and

transmit the second question output by the second machine-learning model over the network to a terminal of the medical doctor in response to an instruction received from the terminal of the patient, the instruction being received based on input by the patient to the user interface, after the second question is displayed on the terminal of the patient and approved by the patient, so that the patient receives an answer related to the medical treatment of the patient prior to performance of the medical treatment of the patient.