Illicit Account Detection

Illicit account detection aims to identify accounts used for fraudulent or criminal activities across various online platforms, including financial networks and messaging apps. Current research heavily utilizes graph neural networks, often incorporating multimodal data (transaction details, network structure, textual information) and self-supervised learning techniques to improve accuracy and efficiency in detecting these accounts. These advanced models offer significant improvements over previous methods, achieving higher detection rates and providing valuable insights for investigators by leveraging explainable AI approaches. The impact of this research is substantial, enabling more effective prevention and mitigation of financial cybercrime and other online illicit activities.

Papers

October 20, 2023

Enhancing Illicit Activity Detection using XAI: A Multimodal Graph-LLM Framework
Jack Nicholls, Aditya Kuppa, Nhien-An Le-Khac
Deep Learning Model xAI Community Financial Fraud Detection Multi Modal LLM Novel Multimodal Illicit Account Detection

September 4, 2023

Effective Multi-Graph Neural Networks for Illicit Account Detection on Cryptocurrency Transaction Networks
Zhihao Ding, Jieming Shi, Qing Li, Jiannong Cao
Malicious Node Multi Graph Neural Network Illicit Account Detection Cryptocurrency Transaction Network

December 6, 2022

Self-supervised Graph Representation Learning for Black Market Account Detection
Zequan Xu, Lianyun Li, Hui Li, Qihang Sun, Shaofeng Hu, Rongrong Ji
Graph Neural Network Self Supervised Learning Weibo Dataset Self Supervised Graph Representation Subgraph Representation Learning Illicit Account Detection

Illicit Account Detection

Papers

Enhancing Illicit Activity Detection using XAI: A Multimodal Graph-LLM Framework

Effective Multi-Graph Neural Networks for Illicit Account Detection on Cryptocurrency Transaction Networks

Self-supervised Graph Representation Learning for Black Market Account Detection