Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection
Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection
复制标题
DOI：
10.1145/3397271.3401253
复制
发表时间：
2020-05
期刊：
影响因子：
--
通讯作者：
Zhiwei Liu;Yingtong Dou;Philip S. Yu;Yutong Deng;Hao Peng-
中科院分区：
文献类型：
--
作者：
Zhiwei Liu;Yingtong Dou;Philip S. Yu;Yutong Deng;Hao Peng-
基于图的模型已广泛用于欺诈检测任务。由于图神经网络（GNN）的发展，最近的工作提出了许多基于同构图或异构图的基于 GNN 的欺诈检测器。这些工作利用现有的 GNN 并聚合邻居信息来学习节点嵌入，这依赖于邻居共享相似上下文、特征和关系的假设。然而，欺诈者带来的不一致问题却很少被研究，即上下文不一致、特征不一致和关系不一致。在本文中，我们介绍了这些不一致之处，并设计了一个新的 GNN 框架 GraphConsis 来解决不一致问题：（1）对于上下文不一致，我们建议将上下文嵌入与节点特征相结合； （2）针对特征不一致，设计一致性得分来过滤不一致的邻居并生成相应的采样概率； （3）对于关系不一致，我们学习与采样节点相关的关系注意权重。对四个数据集的实证分析表明，不一致问题在欺诈检测任务中至关重要。大量实验证明了 GraphConsis 的有效性。我们还发布了一个基于 GNN 的欺诈检测工具箱，其中包含 SOTA 模型的实现。代码可在 \urlhttps://github.com/safe-graph/DGFraud 获取
Graph-based models have been widely used to fraud detection tasks. Owing to the development of Graph Neural Networks~(GNNs), recent works have proposed many GNN-based fraud detectors based on either homogeneous or heterogeneous graphs. These works leverage existing GNNs and aggregate the neighborhood information to learn the node embeddings, which relies on the assumption that the neighbors share similar context, features, and relations. However, the inconsistency problem incurred by fraudsters is hardly investigated, i.e., the context inconsistency, feature inconsistency, and relation inconsistency. In this paper, we introduce these inconsistencies and design a new GNN framework, GraphConsis, to tackle the inconsistency problem: (1) for the context inconsistency, we propose to combine the context embeddings with node features; (2) for the feature inconsistency, we design a consistency score to filter the inconsistent neighbors and generate corresponding sampling probability; (3) for the relation inconsistency, we learn the relation attention weights associated with the sampled nodes. Empirical analysis on four datasets demonstrates that the inconsistency problem is critical in fraud detection tasks. Extensive experiments show the effectiveness of GraphConsis. We also released a GNN-based fraud detection toolbox with implementations of SOTA models. The code is available at \urlhttps://github.com/safe-graph/DGFraud