Based Text Filtering Isolation Technology

Gatekeeper is a new network security technology, which serves as a security firewall between networks. With the wild use of Internet in government and business, the security of network has become the focus. Gatekeeper is a combination of hardware and software systems. This paper is to research the text filtering technology, especially the filtering algorithm based on the vector space and the text in the form of vector space, which enables us to improve the performance of security filtering in gatekeeper technology by choosing text filter.Thesis introduces the theory of gatekeeper and information processing and other related knowledge, and proposes a guideline of text information filter that applies to the gatekeeper aiming at text characteristic selection, filtering algorithm, text information filtering model and text pre-processing technology. the design of plan is Suitable for text information filtering of Gatekeeper, to enhance existing Gatekeeper filtration performance of the system modules, And feature extraction system, the key to hybrid technology and improved algorithm for KNN filter-depth study.By using the traditional characteristic section and the rough theory the combination of hybrid feature extraction method, First, the use of traditional feature selection methods as an initial choice of features , And then ,through the rough set feature selection greatly reduced the number of features to deleting redundant attributes, And a significant reduction in the dimensions of the text, Through the conventional KNN filtering algorithms and the combination of text clustering techniques to improved KNN algorithm , Results showthe improved KNN algorithm not only shortened the processing time also increased the effect of filtering, so improved KNN algorithm apples to gatekeeper...