Latent Semantic Analysis-based Spam Filtering System Design And Realization

With the development and popularization of Internet, E-mail became widely used. Apart from the convenient it bought to people. But one new problem which has arisen is the emergence of junk mails. These junk mails not only consume great amount of web resources, but also spread bad information which do a lot of harm to the society. So the study of how to filtrate those junks has great significance.At first, the paper has introduced the definition of and the actuality of junk mail. Secondly, Latent Semantic Index (LSI) has been analyzed and introduced on detail in this paper. Including basic theory about Latent Semantic Index (LSI) and application field for LSI. On these bases, a mail filtration model is proposed with the main functional module based on LSI to analyze and filtrate E-mails efficiently. The general structure of the E-mail filtration model has described with the introduction of the component modules and functions of sub-modules. The design of various modules is defined in detail. At last, this paper has introduced the test environment, test methods and test results statistic of mail filtration system.The final production of the study is the design and implementation of an E-mail filtration system which based on Latent Semantic Index. With this system the content of E-mail can be analyzed quickly and then junk mail can be interdicted simultaneously. The system had already been put into use and showed good result.