缓解图神经网络应用于欺诈检测的不一致性问题
Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection
摘要：
基于图的模型可以帮助在线检测可疑欺诈。由于图神经网络~（GNNs）的发展，已有的研究工作提出了许多基于同构图或异构图的基于GNNs的欺诈检测框架。这些工作遵循现有的GNN框架，通过聚合邻居信息来学习节点嵌入，这是基于邻居共享相似的上下文、特征和关系的假设。然而，不一致性问题几乎没有被研究，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性，并设计了一个新的GNN框架，$\mathsf{GraphConsis}$，以解决不一致性问题：（1）对于上下文不一致性，提出将上下文嵌入与节点特征相结合;（2）对于特征不一致性，设计一个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）对于关系不一致性，我们学习与采样节点相关联的关系注意力权重。对四个数据集的实证分析表明，不一致性问题是欺诈检测任务中的关键。大量的实验证明了$\mathsf{GraphConsis}$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。该代码可在https://github.com/safe-graph/DGFraud上获得。
英文摘要：
The graph-based model can help to detect suspicious fraud online. Owing to the development of Graph Neural Networks~(GNNs), prior research work has proposed many GNN-based fraud detection frameworks based on either homogeneous graphs or heterogeneous graphs. These work follow the existing GNN framework by aggregating the neighboring information to learn the node embedding, which lays on the assumption that the neighbors share similar context, features, and relations. However, the inconsistency problem 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, $\mathsf{GraphConsis}$, 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, and (3) for the relation inconsistency, we learn a relation attention weights associated with the sampled nodes. Empirical analysis on four datasets indicates the inconsistency problem is crucial in a fraud detection task. The extensive experiments prove the effectiveness of $\mathsf{GraphConsis}$. We also released a GNN-based fraud detection toolbox with implementations of SOTA models. The code is available at https://github.com/safe-graph/DGFraud.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
