This dissertation introduces the conception, framework, model, method and technique of the early warning system. Then a close study is made on the threat assessment system that is subsystem of the early warning system. It draws a good model of the threat assessment and thorough studies the modules of the system. On this condition, the dissertation emphasizes the study on some key techniques of the threat assessment system. At first, utilizing the mending of the Apriori arithmetic, it can enhance efficiency of the mining association rules in threat database; secondly, a good data fusion arithmetic is found through the contrasting the advantages and disadvantages of the different data fusion arithmetic, and a instance of this arithmetic represents how it can be used; finally, the dissertation draws a new arithm(?)tic which can be used measuring the attack or threat in the network and gives a example for the using of this arithmetic. |