Understanding the
user's access pattern is a prerequisite for maintenance and upgrade of enterprise information system. In order to investigate the burstiness, memory, and periodicity characteristics of user's access to information systems, human dynamics was used to analyze 360 631 access records of 314 users in a shipbuilding manufacturing enterprise. The results show that the interval time distribution of user's access to enterprise information systems follows a power-law distribution. The exponents of general users, managers, and users are different. The results also indicate that the behaviors of these three types of user's access show strong burstiness and weak memory, and the user's access to information systems has obvious periodicity.