Clustering Optimization and Evaluation of Campus Network User Behavior Analysis System
The access logs of the flow control server in the campus network of A university are extracted and analyzed in this paper. A hybrid clustering combined with sampling, K-means algorithm and agglomerative hierarchical method is proposed to analyze users’ behavior and classify users’ access objectives and habits, which can not only make clustering results more stable, but also enhance the analysis efficiency of the algorithm.
KeywordsUser behavior analysis Campus network Data mining Cluster analysis
- 2.Kimberly, Y.: The evolution of Internet addiction disorder, Internet addiction. In: Studies in Neuroscience, Psychology and Behavioral Economics, pp. 3–18. Springer, Cham (2017)Google Scholar
- 5.Kate, T.: Social (network) psychology: how networks shape performance, persistence, and access to information. ProQuest Dissertations Publishing, Columbia University (2019)Google Scholar
- 6.Xuxiao, G., Qun, S., Zhongnan, F., Jie, Q.: Multi-dimensional behavior analysis and optimization of traffic in campus network. J. Huazhong Univ. Sci. Technol. (Nat. Sci. Ed.) 44(z1), 131–137 (2016)Google Scholar
- 7.Ling, W., Jiagui, Y., Jinsong, C., Pingshui, W.: Research and implementation of VPN access log analysis platform based on Hadoop. J. Shenyang Univ. (Nat. Sci.) 28(6), 488–496 (2016)Google Scholar
- 8.Zhengguo, Z.: Analysis and research of campus network users’ behavior based on K-means algorithm. J. Anqing Teachers Coll. (Nat. Sci. Ed.) 23(1), 53–56 (2017)Google Scholar