Abnormal Detection of User Behavior in Online Banking
Abstract— Abnormal detection is very important in online banking security. One of the most difficult issues
in abnormal detection is how to calculate the distance between data samples. After the analysis of user
behavior of online banking, we propose a mixed method based on Euclidean distance and cosine similarity,
to measure the similarity among user behaviors. This paper develops an approach to catch similar behaviors
to the abnormal behaviors in online banking transactions, by using the mixed similarity measurement.
Experiment results show that our method can improve the performance of abnormal detection on the
underground dataset, comparing to Euclidean distance and cosine similarity.
Index Terms— Online banking; user behavior; abnormal detection
Yuan Wang, Liming Wang, Wei An
Institute of Information Engineering, Chinese Academy of Sciences, Beijing, CHINA
School of Cyber Security, University of Chinese Academy of Sciences, CHINA
Cite: Yuan Wang, Liming Wang, Wei An, "Abnormal Detection of User Behavior in Online Banking," Proceedings of 2019 the 9th International Workshop on Computer Science and Engineering, pp. 551-557, Hong Kong, 15-17 June, 2019.