Abstract—Traditional authentication methods based on user browsing behaviors consider relatively one-snidely on user browsing habits. They mainly research on the relationships between the sequences of websites or contents without considering user habits comprehensively. So the accuracy when they distinguish different users’ web browsing behaviors cannot ensure enough safety, which can be further optimized. This paper introduces a new method which studies from favorite websites, contents and periods of browsing time. It uses Apriori algorithm to mine user’s frequent itemsets along with the text classification method and normal distribution to calculate access periods of time. Logic regression algorithm is applied onto user authentication. Experiment shows that detection rate can reach 92.7% while false alarm rate is 6.4%.
Index Terms—Identity authentication, user behavior, web browsing features, frequent itemset.
M. Q. Ji, P. H. Zhao, M. M. Wang, C. G. Yan, Z. Li and C. J. Jiang are with the Department of Computer Science and Technology, the Key Laboratory of Embedded System and Service Computing, Ministry of Education, Tongji University, Shanghai, China, 200092 (e-mail: email@example.com, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org).
Cite: Mengqing Ji, Peihai Zhao, Mimi Wang, Chungang Yan, Zhong Li, and Changjun Jiang, "A Kind of Authentication Method based on User Web Browsing Features," International Journal of Machine Learning and Computing vol. 7, no. 2, pp. 18-23, 2017.