Campus Network User Behavior Analysis Based On Mass Access Log

With the deepening of China’s education information construction, computer and network technology are increasingly being used in Education, the campus network came into being. However, with the increasing number of users of the campus network, the management is more and more difficult, at the same time some problems are also exposed. In this case, the research of users’ online behaviors of the campus network is particularly important for network managers.This thesis will be with the campus network of Dalian Maritime University as an example, combining data mining technology with the campus network management, and design to obtain users’ behaviors of the campus network from vast amounts of log data. In view of the current problems, analyzing users’ online behavior of the campus network can provide a scientific basis for the administration and network management, in order to standardize the management, rationally allocate network bandwidth, enhance information security and ensure the stable and efficiency of the campus network environment.The main content is as follows:First, the users’access log data is massive, if we analyze the original data directly, it is needed to pay a higher time and space costs. Based on the characteristics of users’ data, data compression is used to improve the efficiency of analysis, as a precondition that the loss of information is within the allowable range.Then, by statistically analyzing the distribution characteristics of the users’ access connect behavioral variables over time, the law of the frequency of variables and the correlations between variables, basic distribution characteristics of the data is understand, then based on the results of the statistical analysis, the clustering method is used to further explore the specific mode of the user’s online behavior.Moreover, different types of users is significantly different in online time, that is also as a standard to measure users’ online habits in time, the clustering method is used to group the users who have different online habits in time based on weekday and weekend data respectively, and then manage the different groups differently. Finally, according to the results of the experiment, we can find the problems of the campus network and analyze the causes of the problems, and ultimately table a optimized proposal.