O2O模式下货运平台防欺诈系统的设计与应用

Design and Implementation of Users Fraud Prevention System for Freight Platforms in O2O Mode

用户利用O2O模式下货运平台交易成本低、交易速度快、交易灵活性高的特点而套取货运平台返利的欺诈行为,是制约货运平台健康发展的一大因素。本文基于O2O模式下货运平台的实际业务背景和实际需求,设计了一种实现自动化、智能化预防及检测用户欺诈行为的解决方案——用户防欺诈系统。该系统能够与货运平台其他功能系统无缝对接,从返利的发放、货主下单到司机完成订单等业务环节对用户行为进行自动化、智能化管理。通过在Z平台等货运平台的应用实践,得出了在当前线上货运市场真假单比例为1:20的情况下,该系统对异常订单的判断准确率达到80%,对正常订单的判断准确率达到99%,查全率达到62%的结论,在人工检测的基础上,能够提升检测效率164.07%。

It is a big factor that restricts the normal development of freight platform that users make use of the characteristics of the freight platform in O2O mode,which includes low transaction cost,fast transaction speed and high trading flexibility,to take advantage of the platform.Based on the actual business background and actual demand of the freight platform in O2O mode,a brand new intelligent,automated prevention and detection of user fraud framework,entitled 'users fraud prevention system,UFPS'was designed,which could integrate other systems of the freight platform to achieve automatic and intelligent management in coupons,placing order and achieving order.At last,by the practical application in platform Z,it is concluded that in the current online freight market,which has a real and false proportion of 1:20,the accuracy rate of judging the abnormal order is 80%,the accuracy rate of judging the normal order is 99%,and the recall is 62%.What’s more,on the basis of manual detection,the detection efficiency can be improved by 164.07%.