Identity Recognition Based On The Hierarchical Behavior Characteristics Of Network Users

The core idea of ensuring network system security is identity recognition of the user. However, how to identify hackers after breaking through the existing system access control mechanism is still an important problem to be resolved. Therefore, this paper proposes an identity recognition method based on hierarchical behavior characteristics of network users. Behavior of network user was divided into interactive behavior characteristic and mouse behavior characteristic. After characteristics fusion, the Random Forest (RF) method was used to construct the user's identification model. And the identification results of single level behavior characteristics were compared with the results of this paper. The results show that the average True Acceptance Rate (TAR) and False Acceptance Rate (FAR) of 8 users' identity recognition were 82.73% and 7.26%, respectively, which is better than the identification result of single level behavior characteristics. This study provides a new idea for identity recognition based on user behavior. Combining the user's macro interaction behavior characteristics and micro mouse behavior characteristics in a short time or with a small amount of data can better identify users. This adds a layer of security protection for network security.