The Architecture And Solution Design Of Internet Real-time Risk Control About P Company

Internet payment has become a necessary part in our life,but it isn't totally safe.The user's payment experience has bean badly influenced because many risk incidents occur frequently and more and more seriously.P Company as a big-scale third-party payment company has gained some experience in the field of risk control.The system of risk control is mainly based on the post risk control,which can stop the risksat the time of fund settlement or remittance and can prevent most of the risks happening.However,the system is not an overall system that has disadvantages,such as the bad experience from customers,the weak intensity of risk control,which can easily give chances to the crimes.This thesis is based on the theory of real-time risk control,which carries on the architecture and solution design of the real-time risk control and applies in the present system of risk control in P company by using the data mining algorithmand the cloud computing technique.First of all,the thesis demonstrates the disadvantages of the current system of P Company from the theory of real-time risk management,and raises an overall risk control system that majors in real-time risk control and uses quasi-realtime risk control and post risk control as subsidiary roles combining with the resent risk situation in P company.Here,the types of real-time risk are divided into merchant risk control,order risk control,user risk control and transaction risk control.The speed of real-time risk control can be accelerated by putting the work that will take a long time into the quasi-realtime risk control.It classifies the risk levels of merchant,user and transaction by applying data mining algorithm to predict the future risk levels.The risk levels can be applied in the control modules of rules.Secondly,in the application architecture,the thesis divides the real-time risk control system into eight subsystems,such as real-time risk control SOA subsystem,rule engine subsystem,quasi-realtime risk control subsystem,data collecting subsystem,data warehouse subsystem,risk rating subsystem,risk monitor subsystem and risk operation subsystem.All these subsystems cooperate and consist of the system of real-time risk control.In the aspect of technology,the thesis studies cloud computing and renews the design of data access architecture,SOA architecture,message queuing architecture and automatic expansion system.In the aspect of network architecture,the thesis uses the IDC solution plan with two places and four data centers and uses virtual technology in application deployment.In the aspect of database architecture,the thesis carries the horizontal and vertical split towards database.In the aspect of performance testing architecture,it comes up with the single-scene stresstest and full-scene stability test,which requires the product target of real-time risk SOA interface,to meet 720TPS/0.2s and 7*24 hours stability test requirements.All these technological measures increase the real-time ability and reliability as well as give a strong technology support to the real-time risk control.The thesis also studies the strategic management and the operation management of real-time risk control.It analyzes the advantages,disadvantages,chances and threatens of the system of real-time risk control in P company by using SWOT method and studies the establishment of operation system and the building of operation team by using 4P method.Finally,the thesis describes the real risk result of real-time risk control and analyzes the flaws and the improving suggestions for the new system.