基于深度学习的广域网教育信息安全访问自动化控制系统设计  

Design of a wide area network education information security access automation control system based on deep learning

作　　者：宁梓淯[1] 李萌[1] 李燕[1] 高宇 任重贵[1] NING Ziyu;LI Meng;LI Yan;GAO Yu;REN Zhonggui(Daqing Campus,Harbin Medical University,Daqing 163319,China)

机构地区：[1]哈尔滨医科大学大庆校区,黑龙江大庆163319

出　　处：《电子设计工程》2025年第6期100-103,108,共5页Electronic Design Engineering

摘　　要：为了提升教育信息安全,设计基于深度学习的广域网教育信息安全访问自动化控制系统。利用级联多任务卷积神经网络,判定用户身份与用户访问行为是否安全,依据深度学习引擎输出结果,确定用户访问行为的信任值,监督管理用户的身份安全。判定访问控制引擎对用户访问行为的权限鉴定结果,用户未通过权限鉴定时,自动拒绝用户访问请求,断开用户终端的访问请求;用户通过权限鉴定时,将用户访问转移至所需的教育信息资源,实现教育信息访问的自动化控制。测试结果表明,该系统可以精准判定用户的非法访问、异常登录、越权访问等行为,自动中断用户与教育信息的连接。In order to improve the security of educational information,an automatic access control system for wide area network educational information based on deep learning is designed.The cascade multi-task convolutional neural network is used to judge whether the user identity and user access behavior are safe,and the trust value of user access behavior is determined according to the output result of deep learning engine,so as to supervise and manage the user identity safety.Judging the authority authentication result of the access control engine on the user’s access behavior,and automatically rejecting the user’s access request and disconnecting the user terminal’s access request when the user fails the authority authentication.When the user passes the authority authentication,the user’s access will be transferred to the required educational information resources to realize the automatic control of educational information access.The test results show that the system can accurately determine users’illegal access,abnormal login,unauthorized access and other behaviors,and automatically disconnect users from educational information.

关 键 词：深度学习 广域网 教育信息 安全访问 自动化 控制系统 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN03[自动化与计算机技术—计算机科学与技术]