The Research And Application Of Knowledge Discorvery In Database Of Freight Agent Management System

Combining with the using of Freight Agent Management System ,which developed by SunnySoft (Dalian) Co,Ltd , this paper studied the process of discovering the hiden and unknown informations by the advanced technology of KDD.This paper contained five parts. The first part introduced the problem's origin , status que of inside and outside. The second part described the Freight Agent Management System.and pictured the main flow of the Marine Freight Agent Management in detail ,and let us defmitude the position of Decision-making sub-system in it. By explaining the dataflows ,control flows and all modules ,we can outline the KDD. The third part introduced the KDD in the artificail intelligence.lt anatomied the conception ,content .essence and core of KDD,then brought forward the methodology ofSAS------SEMMA. The forth part , the main part of the paper.described all the processof KDD------data preparation ,then data mining according to the SEMMA. This part putforward the system conception of KDD and the Apriori algorithm.Then evolved the create-frequent-set algorithm which was fit for the Freight Agent Management System.Because of the shortage of efficiency ,1 improved the algorithm.Because some of the items were not boolean variables ,1 need the quantitaitve attributes association rules discovering algorithm. In general ,there had the levels among the items,so multilevel association rules existed.After perfecting the algorithmic need interpret and evaluate the knowledge. In the end ,1 discussed the privacy and security of KDD.The fifth part described the future problems and prospect.It improved that KDD was useful to the freight agent department and made benefit.