电信反欺诈系统(AFS)的设计与实现

The design and implementation of an anti-fraud system of telecom

AFS系统(Anti-FraudSystem)通过运用数据挖掘技术中的贝叶斯分类器(BayesianClassifiers)来解决电信行业中的客户欺诈问题。首先介绍了开发电信反欺诈系统的必要性和紧迫性,论述了采用贝叶斯分类器的理由,然后介绍了AFS系统的构架和功能,并着重阐述了系统中采用贝叶斯分类器实现反欺诈功能的具体思路,最后对AFS系统进行了总体评价。

An anti-fraud system is designed in this paper using the Bayesian classifiers of data mining technology to solve the client-fraud problems in telecom trade. The paper firstly expresses the necessary and pressure of designing and implementing the anti-fraud system of telecom, and then discusses the reason of using the Bayesian classifiers of data mining technology in AFS. The paper then introduces the framework and functions of AFS, and specially describes the implementation of the anti-fraud system based on Bayesian classifiers. At last the performance valuation is given.