Frequent sequence mining from massive access log for user’s behaviour investigation
W. Chen, Y. Tong, J. Zhang, T. Qin
With the fast development of Web 2.0, users can obtain everything that they want from the Web. and their access behaviours are recorded by the access log. Based on mining the frequent access
sequence, we can deeply understand their access interests. In turn, it can improve the efficiency of network management. In this paper, we firstly present the methods for log pre-processing and
extract the features. Secondly, we employ the PrefixSpan algorithm to achieve the goal of frequent sequences mining. In order to process the massive log data in network today, we also combined the proposed methods with Spark. Finally, experimental results based on the log data collected from the campus network of Xi’an Jiaotong University verify the efficiency of the developed methods, which are useful for the understanding and management of the user’s behaviour.