A user information inferring system defines user information according to five categories, that is, user environment, circumstance, condition, request, and action, and further defines a sixth category, that is, situation data encompassing these other five categories. User information in each of the defined categories, that is, environment, circumstance, condition, request, action, and situation, is then described using item values, and operating units infer the item values. The inference operation uses temporal-spatial characteristics indicative of the effect of temporal-spatial elements on the item values, information about inferred situations, a rule database containing rules describing the relationships between items, and an optimization process for the defined categories using vector expressions.