Disclosed is a lifestyle pattern classification system for providing a lifestyle pattern expression corresponding to various ways in which a user spends a day according to sensor data and means for classifying a person thereby. From sensor data recording a lifestyle of a user, daily cluster groups representing usual days are extracted by clustering, lifestyle patterns are represented by converting emerging patterns of the daily clusters into feature values, and users are clustered centering around lifestyle patterns extracted thus to generate user classifications.
