Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (8): 2634-2642.DOI: 10.11772/j.issn.1001-9081.2023081153

• Frontier and comprehensive applications • Previous Articles     Next Articles

Credit card fraud detection model based on graph attention Transformation neural network

Fan YANG1, Yao ZOU2,3, Mingzhi ZHU2,3, Zhenwei MA1, Dawei CHENG2,3(), Changjun JIANG2,3   

  1. 1.China UnionPay,Shanghai 200135,China
    2.Department of Computer Science and Technology,Tongji University,Shanghai 201804,China
    3.Collaborative Innovation Center for Internet Financial Security,Tongji University,Shanghai 201804,China
  • Received:2023-08-28 Revised:2023-09-20 Accepted:2023-10-08 Online:2024-08-22 Published:2024-08-10
  • Contact: Dawei CHENG
  • About author:YANG Fan, born in 1982, M. S., senior engineer. His research interests include data mining, financial risk prevention and control, financial big data.
    ZOU Yao, born in 2001, M. S. candidate. Her research interests include financial big data, graph neural network, financial transaction risk prevention and control.
    ZHU Mingzhi, born in 2000, M. S. candidate. His research interests include financial data security, graph neural network, big data analysis and mining.
    MA Zhenwei, born in 1993, Ph. D., senior engineer. His research interests include data mining.
    JIANG Changjun, born in 1962, Ph. D., professor,Academician of the Chinese Academy of Engineering. His research interests include software and information services, network financial security, network transaction risk prevention and control.
  • Supported by:
    National Natural Science Foundation of China(62102287);Shanghai Science and Technology Innovation Action Plan Project(22511100700)

基于图注意力Transformer神经网络的信用卡欺诈检测模型

杨帆1, 邹窈2,3, 朱明志2,3, 马振伟1, 程大伟2,3(), 蒋昌俊2,3   

  1. 1.中国银联,上海 200135
    2.同济大学 计算机科学与技术系,上海 201804
    3.同济大学 网络金融安全国家级协同创新中心,上海 201804
  • 通讯作者: 程大伟
  • 作者简介:杨帆(1982—),男,重庆人,高级工程师,硕士,主要研究方向:数据挖掘、金融风险防控、金融大数据
    邹窈(2001—),女,湖南邵阳人,硕士研究生,主要研究方向:金融大数据、图神经网络、金融交易风险防控
    朱明志(2000—),男,上海人,硕士研究生,主要研究方向:金融数据安全、图神经网络、大数据分析与挖掘
    马振伟(1993—),男,安徽阜阳人,高级工程师,博士,主要研究方向:数据挖掘
    程大伟(1987—),男,江苏淮安人,副教授,博士,主要研究方向:金融大数据、大数据分析与挖掘、图神经网络、强化学习 dcheng@tongji.edu.cn
    蒋昌俊(1962—),男,安徽安庆人,教授,中国工程院院士,博士,主要研究方向:软件与信息服务、网络金融安全、网络交易风险防控。
  • 基金资助:
    国家自然科学基金资助项目(62102287);上海市科技创新行动计划项目(22511100700)

Abstract:

For the issue of existing models’ inability to accurately identify intricate and diverse patterns of gang fraud, a new practical credit card fraud detection model based on complex transaction graph was proposed. Firstly, the association transaction graph was constructed based on the original transaction information of the users, then the graph Transformer neural network module was employed to mine the gang fraud characteristics directly from the transaction network without cumbersome feature engineering. Finally, the high-precision detection of fraud transactions was realized by jointly optimizing the topological features and sequential transaction features by the fraud detection network. The credit card anti-fraud experiment results showed that the proposed model outperformed seven benchmark models in all evaluation indexes. The Average-Precision (AP) improved by 20% and the Area Under the ROC Curve (AUC) increased by an average of 2.7% over the best benchmark Graph Attention Network (GAT) model in transaction fraud detection tasks. These results indicate that the proposed model is effective in the detection of credit card fraud transactions.

Key words: credit card transaction, fraud detection, graph neural network, self-attention Transformer, heterogeneous graph

摘要:

针对现有模型无法精准识别复杂多变的团伙诈骗模式的问题,提出一种新型实用的基于复杂交易图谱的信用卡反欺诈检测模型。首先,利用用户原始的交易信息构造关联交易图谱;随后,使用图自注意力Transformer神经网络模块直接从交易网络中挖掘团伙欺诈特征,无需构建繁冗的特征工程;最后,通过欺诈预测网络联合优化图谱中的拓扑模式和时序交易模式,实现对欺诈交易的高精度检测。在信用卡交易数据上的反欺诈实验结果表明,所提模型在全部评价指标上均优于7个对比的基线模型:在交易欺诈检测任务中,平均精度(AP)比基准图注意力神经网络(GAT)提升了20%,ROC曲线下方面积(AUC)平均提升了2.7%,验证了所提模型在信用卡欺诈交易检测中的有效性。

关键词: 信用卡交易, 欺诈检测, 图神经网络, 自注意力Transformer, 异构图

CLC Number: