Scholarlink

In various domains, such as social and computer networks, representing complex interconnections between entities is increasingly critical. Traditional methods often fail to capture these intricacies, leading to the adoption of graph-based models. This work focuses on anomaly detection in Knowledge Graphs (KGs) using advanced Graph Neural Network (GNN) architectures. We explore KG design techniques, enhance GNN models with MLP layers, and address class imbalance through oversampling. Empirical evaluations on real-world datasets demonstrate improved efficiency and accuracy in anomaly detection, offering a robust framework for applications in cybersecurity, network management, and social network analysis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Efficient Graph Construction and Neural Networks for Anti-money Laundering

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • An Tran,
  • Hoang Ngan Nguyen,
  • Khang Nguyen,
  • Anh Viet Nguyen,
  • Nga Mai

摘要

In various domains, such as social and computer networks, representing complex interconnections between entities is increasingly critical. Traditional methods often fail to capture these intricacies, leading to the adoption of graph-based models. This work focuses on anomaly detection in Knowledge Graphs (KGs) using advanced Graph Neural Network (GNN) architectures. We explore KG design techniques, enhance GNN models with MLP layers, and address class imbalance through oversampling. Empirical evaluations on real-world datasets demonstrate improved efficiency and accuracy in anomaly detection, offering a robust framework for applications in cybersecurity, network management, and social network analysis.