The Time Statistic Feature Of Users’ Access Behavior In EIS And The Analysis Of Its Dynamics

Over the past three decades, information systems play a great role in enhancing operational efficiency of the enterprises, improving the management and decision-making,and strengthening the enterprises’ competitiveness. Some of them have even become the strategic weapon for enterprises to participate in market competition. Nowadays, the emergence of new information technologies, such as cloud computing, Internet of Things,mobile commerce, makes the global data explode tremendously, which means human society has turned into the Big Data Era. In the meantime, this trend offers a golden opportunity for enterprises to upgrade their information systems and create new competitive advantages.Accurate mastery of the accessing regularity and the actual needs of the users is the prerequisite for the optimization and upgrading of the enterprise information systems through utilizing burgeoning information technology. This study conducts a quantitative analysis on the laws of users’ behavior according to the data of user access logs of management information system provided by certain enterprise. The study does a statistical analysis on the interval time distribution of the users’ access to information systems and some indexes of the liveness, power exponents, burstiness and memory indicators. The periodicity of user’s access behavior in information systems is investigated from two angles, that is, the number and character of users. The result shows that users’ access behavior in information systems presents the feature of serious fat tail, as well as the character of non-Poisson and obvious periodicity of fluctuation. However, this can not be described by a single distribution. The research emphases mainly lies in the following points:(1) The study of the laws of users’ access behavior in enterprise information system based on the number of users. The research firstly explores the laws of users’ access behavior in enterprise information system. The act proceeds from three levels, namely,individual user, group users and all users. Meanwhile, the study probes into the relation between the activity and the power exponent of individual user’s access behavior. The research results show that the three aspects present intricate statistical properties. Individual user’s access to enterprise information system displays strong paroxysm and weak memorability. The activity and the power exponent form a positive correlation, that is to say, the power exponent will increase along with the rise of activity. But the powerexponent will not increase without limit, and it would remain stable when reaching about1.5.(2) In view of the impact of the users’ jobs on the access of enterprise information system, the research classifies the users into three categories, that is, normal users,managers, and system administrators, in the hope of confirming the access behavior of different category of users when they visit the management information system in enterprise. The study respectively discusses the statistical properties of the time when the three categories of users access to the management information system.(3) An old saying goes that, work begins at sunrise and stops at sunset. Now we have nine-to five jobs. All these represent the obvious periodicity of human’s daily behavior. The research also explores the periodicity of user’s access behavior in enterprise information system. The findings show that the user’s access to enterprise information system presents strong fluctuations and periodicity measured in weeks, whereby comes up with corresponding strategies of the operation and maintenance of the information systems.