Based On The Classification Recognition Of Business Early Warning Application Control

Over the last decade, China’s trade administration has been considerably developedin reform and progress, but some control technologies in management sometimes lagbehind what is actually happening. When a trade risk such as quality problems mayoccur but you are not aware of the risk and do not have early warning, the risk will incurlosses such as increased cost or bad account.In order to address the above problem, business is classified into two types: whoserisk is controllable and whose risk is uncontrollable; and early warning control isapplied to the business management module. Early warning identification (such as inthe PURCHASE interface) is the core function of the software, for example, if a(language) description in submission is identified by the early warning interface thatquality certification of CCC (CHINA COMPULSORY CERTIFICATE) is required, itwill be held back in the same way because it could not pass the budget cost control.Therefore, a security identification lock for reference is added and interrupts theapproval procedures handled by humans. This is the "central idea" of the software.The model of feature vector is introduced in this thesis. Imagine this case: youhave approved a purchase contract, and the payment for the goods has been effected byLetter of Credit, but the goods are required to have the CCC certificate, so you cannottake delivery of such goods because you don’t have the CCC certificate, then, how tosolve this problem? First of all, L/C should have been rejected as controllable risk in thecontract control. In commercial practices, T/T against copy or original of B/L is usuallyidentified as the character or “quality certificate” clause is added to L/C as the character,but if there is no prior agreement on characters such as “quality certificate clause”,which is often deemed as “a soft clause”, it may result in trade conflicts which mayconstitute a breach of contract.This thesis focuses on the following three main aspects:1) On theory, this thesis elaborates the significance of the establishment of the systemand structure meeting with purchase early warning requirements at the core;2) On application, this thesis proposes new design requirements based on the lack of risk identification in the front-end system, i.e., classification of information in the frontend in order to control risks;3) On technology, this thesis proposes integrated solutions from the viewpoint ofpractical applications, i.e., integration of the topic-related original business informationmodule and a new classification application control module.This thesis emphasizes on the following four key problems against which theabove mentioned solution is proposed on technology:1）the location problem of applying the identification modules2）the building problem of the identification creteria3）the echoing problem of the early warning4）the choosing problem of the identification modulesOn the basis of analyzing Boolean vector model, the thesis describes how to applythe conditional data in the module of conditional base to form dynamic recognition, andthus the system can process by classifying and controlling the modification, submissionand approval of the data of purchasing module. At last, the system will show thewarning color according to the data which don’t meet the criteria, or make frozenprocessing. In the view of the results, the software can realize the early warning inaccordance with the conditions in conditional database, and success to achieve the earlypurchasing warning function at the early position of the business. Therefore, we canmake the conclusion that such early warn design and improvement can control certainbusiness risk and resolve certain business risk problem.