The Security Of E-mail Systems Filter Design

With the INTERNET / INTRANET widely in the enterprise information technology applications, the higher the development of enterprises dependent on the external network information system services, this paper based on network topology, the study of a high-speed Internet information monitoring system, based on the e-mail content filtering system can be achieved right not encrypted strings and accessories in pure compression of pure string without the high-efficiency filtration and screening, and to achieve a very small overhead of system time to better protect against leaks against business interests.As the core of the system components - high-speed Internet information monitoring system, and placed in the network topology in the key positions, the information flow must pass through the Office of the data packets for detailed analysis and monitoring, filtering out the illegal information, in monitoring system, considering the diversity of network protocols and scalability features, and information surveillance system designed to adopt a Plug-in in the form of customization, to provide a unified external interface, so that various protocols can be loaded into the system to, so that no seam connections, thereby greatly enhancing the system versatility and scalability.In order to more fully understand the user's information content, the system filters using a real-time monitoring of the subsystem model - based on the finite automaton multi-mode adaptation algorithm, while achieving high efficiency filtration significantly reduced time and space overhead, its implementation process is: IP data packets directly to the content characteristics of adaptation, feature information to solve the problems brought about miscarriage of justice. Streaming technology can effectively reduce the system load and improve the filtration efficiency, based on the concept through an extension to complete the tree-branch structure of the user's information according to different organizations, in order to achieve a common part of the share, then raised the side of approximation based on the extent and relevance distribution rates, to achieve the standard string and filtering the quantitative fit between the e-mail content.