针对“羊毛党”研发的多防御系统——以反薅盾为例

AbstractAfter 2009,human beings have really entered the era of big data.Thescene of data explosion provides more ways to realize black production,andbusiness security has entered a new stage since then.However,mostenterprises'cognition of black production has not improved with thedevelopment of business,and there are serious information inequalities.If"virtual human"is used to attack enterprises,its harm can be seen.In recentyears,more and more Internet-based marketing and promotion activitieshave achieved good results.Therefore,corresponding measures must be takento effectively identify fraudulent activities.Traditional marketing monitoring and anti fraud platforms usually userelational database and modeling analysis technology to realize themonitoring of marketing activities and the identification of fraud behavior.However,due to the explosive growth of the number of participants inmarketing activities and the increasing demand of users for real-timemarketing activities reward settlement,the traditional platform has graduallyfailed to meet the current needs.Therefore,there is an urgent need for a setof solutions that can effectively prevent and control the "wool party"in themarket.In this paper,taking the anti Huidun anti fraud project that the authorparticipated in and developed personally as an example,a multi effectreal-time business monitoring and anti fraud system including the maindefense system,database,information collection software,intelligentidentification system and intelligent classification system is constructed byusing aut proprietary patent algorithm,delear poker intelligent engine,andcombining with AI main call,data collection,black and white list database,etc.Through experiments and business practice,it is verified that the anti-haodun system is in the anti -wool party business.Key words:wool party anti fraud anti-haodun big data intelligentalgorithm