基于多视角图神经网络的欺诈检测算法

doi:10.11959/j.issn.1000-436x.2022221

通信学报 ›› 2022, Vol. 43 ›› Issue (11): 225-232.doi: 10.11959/j.issn.1000-436x.2022221

基于多视角图神经网络的欺诈检测算法

陈卓, 朱淼, 杜军威

青岛科技大学信息科学技术学院，山东青岛 266061

修回日期:2022-09-30 出版日期:2022-11-25 发布日期:2022-11-01
作者简介:陈卓（1978− ），女，山东青岛人，博士，青岛科技大学副教授、硕士生导师，主要研究方向为自然语言处理、推荐系统等
朱淼（1998− ），女，安徽六安人，青岛科技大学硕士生，主要研究方向为图神经网络、异常检测等
杜军威（1974−），男，山东威海人，博士，青岛科技大学教授、博士生导师，主要研究方向为数据挖掘、知识图谱与知识工程等
基金资助:
国家自然科学基金资助项目(62172249);国家自然科学基金资助项目(61973180);国家自然科学基金资助项目(62202253);山东省自然科学基金资助项目(ZR2021MF092)

Multi-view graph neural network for fraud detection algorithm

Zhuo CHEN, Miao ZHU, Junwei DU

School of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China

Revised:2022-09-30 Online:2022-11-25 Published:2022-11-01
Supported by:
The National Natural Science Foundation of China(62172249);The National Natural Science Foundation of China(61973180);The National Natural Science Foundation of China(62202253);The Natural Science Foundation of Shandong Province(ZR2021MF092)

摘要/Abstract

摘要：

针对欺诈检测领域样本标签不平衡、欺诈节点之间缺乏必要连接，导致欺诈检测任务不符合图神经网络同质性假设的问题，提出了基于多视角图神经网络的欺诈检测（MGFD）算法。首先，利用结构无关的编码器对网络中节点进行属性编码，以学习欺诈节点与正常节点之间的差异，使用层次注意力机制对网络中多视角信息进行融合，在学习差异的基础上充分利用网络中不同视角之间的交互信息对节点进行建模；然后，基于数据不平衡比采样子图，依据欺诈节点连接特性构建样本进行分类学习，解决样本标签不平衡的问题；最后，预测标签判别节点是否为欺诈节点。在公开数据集上的实验表明，MGFD算法在基于图的欺诈检测领域检测效果优于对比方法。

关键词: 欺诈检测, 异常检测, 注意力机制, 图表示学习, 不平衡学习

Abstract:

Aiming at the problem that in the field of fraud detection, imbalance labels and lack of necessary connections between fraud nodes, resulting in fraud detection tasks not conforming to the hypothesis of homogeneity of graph neural networks, multi-view graph neural network for fraud detection (MGFD) algorithm was proposed.First, A structure-independent encoder was used to encode the attributes of nodes in the network to learn the difference between the fraud node and the normal node.The hierarchical attention mechanism was designed to integrate the multi-view information in the network, and made full use of the interaction information between different perspectives in the network to model the nodes on the basis of learning differences.Then, based on the data imbalance ratio sampled subgraph, the sample was constructed according to the connection characteristics of fraud nodes for classification, which solved the problem of imbalance sample labels.Finally, the prediction label was used to identify whether a node is fraudulent.Experiments on real-world datasets have shown that the MGFD algorithm outperforms the comparison method in the field of graph-based fraud detection.

Key words: fraud detection, anomaly detection, attention mechanism, graph representation learning, imbalance learning

中图分类号:

TP183

陈卓, 朱淼, 杜军威. 基于多视角图神经网络的欺诈检测算法[J]. 通信学报, 2022, 43(11): 225-232.

Zhuo CHEN, Miao ZHU, Junwei DU. Multi-view graph neural network for fraud detection algorithm[J]. Journal on Communications, 2022, 43(11): 225-232.

图/表 11

图1

表1

图2

图3

图4

表2

对比实验结果"

方法		Yelpchi			Amazon
方法	AUC	Recallmacro	F1-macro	AUC	Recallmacro	F1-macro
GCN	0.598 3	0.500 0	0.562 0	0.779 4	0.5000	0.648 6
FdGars	0.653 6	0.500 0	0.553 2	0.818 5	0.718 6	0.614 5
GraphConsi	0.698 3	0.610 0	0.585 7	0.874 1	0.851 2	0.751 2
CARE-GNN	0.765 7	0.664 6	0.633 2	0.906 7	0.834 7	0.899 0
FRAUDER	0.772 2	0.677 2	0.591 2	0.925 3	0.881 6	0.866 7
MGFD	$0 . 7910$	$0 . 6781$	$0 . 6541$	$0 . 9256$	$0 . 8910$	$0 . 9241$

表2

图5

图6

图7

表3

表4

参考文献 22

[1]	朱会娟, 陈锦富, 李致远 ,等. 基于多特征自适应融合的区块链异常交易检测方法[J]. 通信学报, 2021,42(5): 41-50.
	ZHU H J , CHEN J F , LI Z Y ,et al. Block-chain abnormal transaction detection method based on adaptive multi-feature fusion[J]. Journal on Communications, 2021,42(5): 41-50.
[2]	POURHABIBI T , ONG K L , KAM B H ,et al. Fraud detection:a systematic literature review of graph-based anomaly detection approaches[J]. Decision Support Systems, 2020,133:113303.
[3]	ZHANG G , WU J , YANG J ,et al. FRAUDRE:fraud detection dual-resistant to graph inconsistency and imbalance[C]// Proceedings of IEEE International Conference on Data Mining. Piscataway:IEEE Press, 2021: 867-876.
[4]	WANG D X , LIN J B , CUI P ,et al. A semi-supervised graph attentive network for financial fraud detection[C]// Proceedings of IEEE International Conference on Data Mining. Piscataway:IEEE Press, 2019: 598-607.
[5]	MCAULEY J J , LESKOVEC J . From amateurs to connoisseurs:modeling the evolution of user expertise through online reviews[C]// Proceedings of the 22nd International Conference on World Wide Web. New York:ACM Press, 2013: 897-908.
[6]	LIU Z , DOU Y , YU P S ,et al. Alleviating the inconsistency problem of applying graph neural network to fraud detection[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York:ACM Press, 2020: 1569-1572.
[7]	DOU Y T , LIU Z W , SUN L ,et al. Enhancing graph neural network-based fraud detectors against camouflaged fraudsters[C]// Proceedings of the 29th ACM International Conference on Information ＆Knowledge Management. New York:ACM Press, 2020: 315-324.
[8]	陈晋音, 张敦杰, 黄国瀚 ,等. 面向图神经网络的对抗攻击与防御综述[J]. 网络与信息安全学报, 2021,7(3): 28.
	CHEN J Y , ZHANG D J , HUANG G H ,et al. Adversarial attack and defense on graph neural networks:a survey[J]. Chinese Journal of Network and Information Security, 2021,7(3): 28.
[9]	MA J , ZHANG D , WANG Y ,et al. GraphRAD:a graph-based risky account detection system[C]// Proceedings of ACM SIGKDD Conference. New York:ACM Press, 2018:9
[10]	BIAN T , XIAO X , XU T ,et al. Rumor detection on social media with bi-directional graph convolutional networks[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto:AAAI Press, 2020: 549-556.
[11]	LI A , QIN Z , LIU R S ,et al. Spam review detection with graph convolutional networks[C]// Proceedings of the 28th ACM International Conference on Information and Knowledge Management. New York:ACM Press, 2019: 2703-2711.
[12]	CHAWLA N V , BOWYER K W , HALL L O ,et al. SMOTE:synthetic minority over-sampling technique[J]. Journal of Artificial Intelligence Research, 2002,16: 321-357.
[13]	WANG W , WANG S , FAN W ,et al. Global-and-local aware data generation for the class imbalance problem[C]// Proceedings of the 2020 SIAM International Conference on Data Mining. Philadelphia:Society for Industrial and Applied Mathematics, 2020: 307-315.
[14]	CHI J F , ZENG G X , ZHONG Q W ,et al. Learning to undersampling for class imbalanced credit risk forecasting[C]// Proceedings of IEEE International Conference on Data Mining. Piscataway:IEEE Press, 2020: 72-81.
[15]	CAO K , WEI C , GAIDON A ,et al. Learning imbalanced datasets with label-distribution-aware margin loss[C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems. Massachusetts:MIT Press, 2019: 1567-1578.
[16]	HU Z , TAN B , SALAKHUTDINOV R R ,et al. Learning data manipulation for augmentation and weighting[C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems. Massachusetts:MIT Press, 2019:32.
[17]	SHI M , TANG Y F , ZHU X Q ,et al. Multi-class imbalanced graph convolutional network learning[C]// Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence. San Francisco:Margan Kaufmann, 2020: 2879-2885.
[18]	ZHAO T X , ZHANG X , WANG S H . GraphSMOTE:imbalanced node classification on graphs with graph neural networks[C]// Proceedings of the 14th ACM International Conference on Web Search and Data Mining. New York:ACM Press, 2021: 833-841.
[19]	TONG H H , FALOUTSOS C , PAN J Y . Fast random walk with restart and its applications[C]// Proceedings of the Sixth International Conference on Data Mining. Piscataway:IEEE Press, 2006: 613-622.
[20]	RAYANA S , AKOGLU L . Collective opinion spam detection:bridging review networks and metadata[C]// Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York:ACM Press, 2015: 985-994.
[21]	WELLING M , KIPF T N . Semi-supervised classification with graph convolutional networks[J]. arXiv Preprint,arXiv:1609.02907, 2016.
[22]	WANG J Y , WEN R , WU C M ,et al. FDGars:fraudster detection via graph convolutional networks in online APP review system[C]// Proceedings of the 22nd International Conference on World Wide Web. New York:ACM Press, 2019: 310-316.

数据集	节点数	不平衡率	关系	关系边数	标签相似度
Yelpchi	45 954	5.9	R-U-R	49 315	0.908 9
			R-S-R	3 402 743	0.176 4
			R-T-R	573 616	0.185 7
Amazon	11 944	13.5	U-P-U	175 608	0.167 3
			U-S-U	3 566 479	0.055 8
			U-V-U	1 036 737	0.053 2

节点	关系	平均余弦相似度
节点	关系	初始嵌入	P₁	P₂
欺诈节点	U-P-U	0.690 7	0.069 0	0.069 1
	U-S-U	0.594 7	0.062 1	0.062 3
	U-V-U	0.507 1	0.063 9	0.064 5
正常节点	U-P-U	0.679 4	0.618 1	0.618 5
	U-S-U	0.594 4	0.533 0	0.562 0
	U-V-U	0.513 1	0.567 1	0.601 2

节点	关系	平均余弦相似度
节点	关系	初始嵌入	P₁	P₂
欺诈节点	R-U-R	0.951 1	0.302 3	0.351 0
	R-T-R	0.871 7	0.101 9	0.101 6
	R-S-R	0.863 4	0.101 9	0.101 9
正常节点	R-U-R	0.981 1	0.865 3	0.865 6
	R-T-R	0.863 6	0.885 2	0.885 3
	R-S-R	0.855 6	0.829 0	0.829 0

基于多视角图神经网络的欺诈检测算法

Multi-view graph neural network for fraud detection algorithm

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 22

相关文章 15

Metrics

推荐阅读 0

[1]	霍纬纲, 梁锐, 李永华. 基于随机Transformer的多维时间序列异常检测模型[J]. 通信学报, 2023, 44(2): 94-103.
[2]	廖建新, 付霄元, 戚琦, 王敬宇, 孙海峰. 6G-ADM：基于知识空间的6G网络管控体系[J]. 通信学报, 2022, 43(6): 3-15.
[3]	段雪源, 付钰, 王坤. 基于VAE-WGAN的多维时间序列异常检测方法[J]. 通信学报, 2022, 43(3): 1-13.
[4]	吴平, 常朝稳, 左志斌, 马莹莹. 基于地址重载的SDN分组转发验证[J]. 通信学报, 2022, 43(3): 88-100.
[5]	冯海林, 张潇, 刘同存. 融合评论文本特征和评分图卷积表示的推荐模型[J]. 通信学报, 2022, 43(3): 164-171.
[6]	孙海丽, 龙翔, 韩兰胜, 黄炎, 李清波. 工业物联网异常检测技术综述[J]. 通信学报, 2022, 43(3): 196-210.
[7]	仲美玉, 吴培良, 窦燕, 刘毅, 孔令富. 基于中文语义-音韵信息的语音识别文本校对模型[J]. 通信学报, 2022, 43(11): 65-79.
[8]	段雪源, 付钰, 王坤, 刘涛涛, 李彬. 基于多尺度特征的网络流量异常检测方法[J]. 通信学报, 2022, 43(10): 65-76.
[9]	王洪雁, 袁海. 基于骨骼及表观特征融合的动作识别方法[J]. 通信学报, 2022, 43(1): 138-148.
[10]	朱会娟, 陈锦富, 李致远, 殷尚男. 基于多特征自适应融合的区块链异常交易检测方法[J]. 通信学报, 2021, 42(5): 41-50.
[11]	赵晓娟, 贾焰, 李爱平, 陈恺. 基于层级注意力机制的链接预测模型研究[J]. 通信学报, 2021, 42(3): 36-44.
[12]	郭璠, 张泳祥, 唐琎, 李伟清. YOLOv3-A：基于注意力机制的交通标志检测网络[J]. 通信学报, 2021, 42(1): 87-99.
[13]	陈铁明,金成强,吕明琪,朱添田. 基于样本增强的网络恶意流量智能检测方法[J]. 通信学报, 2020, 41(6): 128-138.
[14]	戚琦,申润业,王敬宇. GAD：基于拓扑感知的时间序列异常检测[J]. 通信学报, 2020, 41(6): 152-160.
[15]	李琳辉,周彬,连静,周雅夫. 基于社会注意力机制的行人轨迹预测方法研究[J]. 通信学报, 2020, 41(6): 175-183.