At least one feature amount is calculated with sound data uttered by a subject, a degree of a psychosomatic state of the subject is calculated based on the calculated feature amount, and the psychosomatic state of the subject are estimated based on the calculated degree.