The Design And Implementation Of Intrannet Security Control System Based On Machine Learning

VWith the development of computer network technology, bring us convenience, but also produced a lot of new problems. The research focus of the internal network of institutions and enterprises are often the network control and control peripherals such as firewall, secure routing.and intrusion detection, authentication, document security management of network of internal terminal in the file is not deep enough, security problems have become increasingly prominent. Therefore, this thesis focuses on the problem of authentication and file security management in Intranet, and designs and implements the solution for intranet security.Intranet authentication is the basis of realizing intranet security.Under the Windows system, the traditional single factor account password system is not enough to meet the security requirements of Intranet terminals. This paper combines the USBKEY and Windows authentication mechanism, proposed a new authentication solution: from the cache mechanism of Windows system realization of SSO based on USBKEY, using ActiveX technology to realize identity authentication on the Web browser, USBKEY technology to achieve a secure login authentication based on Windows. The implementation of dual factor authentication mechanism based on USBKEY enhances the security of host login.File security management can be divided into two parts: file security storage and file security audit. In this paper, on the basis of TrueCrypt,combined with USBKEY, to achieve a safe encryption algorithm using the specified disk encryption and random security encryption password stored in the USBKEY, effectively solve the security key storage, use disk protection mechanism to prevent user errors on the disk, but also in the disk partition is damaged, by restoring pre encryption the backup mechanism as soon as possible to achieve data. The file flow in the network is convenient, frequent operation, the file security state of concern is difficult to achieve. This paper first through the classification of machine learning classification algorithm SVM the security level of Intranet in the file, the file is divided into ordinary, confidential and top secret files on the three levels of content management; then the operating behavior of file monitoring and reporting, formulate corresponding rules of risk assessment, analysis of the current state of the file server security and can leak early warning for reporting operation events. Through the secure storage of the file security audit effectively implements the file security management within the network.