|
|
|
| Research on the Construction of a "Trinity" University Cybersecurity Defense System in the Context of Artificial Intelligence |
| LI Shuangyuan1，Qing ZheHao2**
|
(1.Information Center,Jilin University of Chemical Technology,Jilin City 132022,China;2.School of Information and Control Engineering,Jilin University of Chemical Technology,Jilin City 132022,China)
|
|
|
|
Abstract With the rapid advancement of artificial intelligence, while enhancing the efficiency of university network security, it has also spawned new types of attacks, leading to a triple contradiction of technological gaps, fragmented management, and ethical conflicts. This paper proposes a "smart perception-dynamic decision-making-hierarchical execution" trinity defense system: the perception layer achieves unsupervised anomaly detection through multi-source data collection and deep clustering; the decision layer generates adaptive strategies with second-level responses based on Markov processes and reinforcement learning; the execution layer employs priority queues to balance security and operational continuity. Experiments demonstrate an anomaly detection accuracy of 94.6%, a decision response time of 1.2 seconds, an execution accuracy of 98.9%, and successful sensitive data desensitization, providing an intelligent foundation for university network security.
|
|
Published: 06 July 2026
|
|
|
|
|
|
|
|
