Study on user’s identification system in insider threats

Abstract

Monitoring user’s abnormal behaviors, which is an effective method to detect impersonation, is used for im-personation detection in insider threats.A model is built by using TAN-based Bayesian network to reflect the characteris-tics of user’s behavior.When the deviation from the model is found, the system can determine the identity of the user.As a result, experiments show that the monitoring numbers of processes called by users can be very effective on detecting impersonation and can identify the identity of the attacker.

Keywords

• insider threat;identification;Bayesian network;extended tree augment naive bayesian