关联关系图谱技术在团伙欺诈识别中的应用

Application of Association Relation Graph Technology in Group Fraud Recognition

摘要: 银行现有的反欺诈体系主要以专家规则为主，在防范新型的欺诈手段和欺诈形势时，存在两大痛点：一是无法识别个体正常但属于团伙欺诈的情形；二是缺乏对风险传导的防控。基于此，本文开展基于关联图谱的反欺诈体系研究，从关联图谱的底层逻辑、构建原理、方法路径以及应用价值等方面阐述关联图谱反欺诈体系的底层逻辑和应用价值，并进一步探索在营销精准化的应用前景。

Abstract: The bank’s existing anti-fraud system is mainly based on expert rules. There are two major pain points in preventing new fraud methods and fraud situations: one is the failure to identify indi-viduals that are normal but belong to group fraud; the other is the lack of risk transmission in prevention and control. Based on this, this article carries out research on the anti-fraud system based on the correlation graph, expounds the underlying logic and application value of the an-ti-fraud system of the correlation graph from the aspects of the underlying logic, construction principle, method path and application value of the correlation graph, and further explores the precision in marketing the application prospects of modernization.