Research And Implementation Of A SAAS-based Ecommerce Fraud Management Platform

Fraud Management refers to the action for checking the risk of cheating and corruption. With setting up the corresponding risk rule and condition, the chance of being cheated or corrupted could be lower down. Or it could reduce the negative impact after the thing really happened. A fraud management system that is well built is helpful for making a secure and harmonious ecommerce environment. With the development of cloud computing technique, building a SaaS based fraud management is possible. And it will make the threshold of introducing the system lower for most companies, and give it the opportunity of growing fast.At the introduction chapter, the article introduces the background, the target and the feasibility of this research. The introduction is started from the origin of the ecommerce, and then describes some concepts of ecommerce. After that, it tells the reality and the trend of ecommerce. Also, it raises the risk and challenge with the growth of ecommerce. Against the risk and challenge, the article brings three targets of the research. They are processing data in a scalable and distributed way, managing fraud in a data driven approach and multi-tenant. And finally, it will bring ecommerce fraud management system that is easy to expand and easy to maintain.At the overview chapter, the article introduces the design model, and describes the easy-to-expand attribute of the system in functionality and performance. It emphasizes the target is an ecommerce fraud management platform that includes fraud management, fraud model analysis and business intelligence. To make the people understand the system design better, the chapter also introduces the concept of fraud management, the concept of SaaS, which also includes the attributes and advantages of SaaS. At the last of this chapter, it describes the different layers of the system and the functionality of each layer. And it also describes the techniques difference of the system.At the architecture chapter, the article introduces the classical 4 plus 1 view of the system. It is started from the system requirement. The article describes the system architecture by using some common UML diagrams, including use case diagram, sequence diagram, activity diagram, package diagram and state diagram. It tells the logical functionality, the concurrence of the processes, the internal implementation and physical structure. For sequence and activity diagram, they mainly describe the do fraud check use case.At the implementation chapter, the article introduces the implementation on database layer, data access layer, business logic layer and frontend layer. It explains the functionality of some key code snippets. This chapter gives the technique answer to how to achieve the three targets of the research. The important techniques includes DAO(data access object), Dynamic Proxy, MyBatis framework and JavaParser tool.At the testing and verification chapter, the article mainly focus on the functionality and performance testing. It introduces the testing strategy and approaches, and then lists the testing execution and results.At the conclusion chapter, the article sums up the achievement of the research. It includes the implementation of processing data in a scalable and distributed way, implementation of managing fraud in a data driven approach and implementation of multi-tenant. In the vision of the future, the article brings some to-be-improved aspects of the system, including generating data analysis reports, build analysis mode with historical data. They make the system keep growing with enough space.