The Reasearch Of The Principle And Application Of Immune Negative Selection Classifier

In this paper, it researches the way of Web text mining by Artificial Immune System.In the first chapter, the background of research is introduced. And then it introduces the development history and the main researches of immunology, because immunology and Artificial Immune System are close related each other.The internal relation is analyzed between them.In the second chapter, the natural immune system is introduced in detail, including the basic components ,mechanisms and principles. The immunology principles on which Artificial Immune System depends mainly include immune network theory, the principle of clonal selection and the principle of negative selection. Artificial Immune System is built on the basis of immunology theory and the mechanisms of immune system.In chapter 3, it focus on the binary model of Artificial Immune System. Most of them are developed for researching the mechanisms of immune system. The later models gradually are transferred to the field of engineering. Farmer's differential equation model, the model based on genes libraries and the co-evolutionary model are mainly discussed in this chapter. The co-evolutionary model is the one that is expanded to Web text mining.In chapter 4, the technologies of Web text mining are discussed in detail. It is a cross field that involves many different technologies field and also include many complex ways of technologies from feature extraction to model and the evaluation ways of models and so on. This chapter and the former three chapters are combined to be the bases of theory and technology of the model of Web text classification given in chapter in chapter 5.In chapter 5, the negative selection model based on co-evolutionary isgiven and compared with the traditional ways. It shows that Artificial Immune System is the feasible way of Web text classification, though there are many shortcomings.