Anomaly Detection in Dynamic Graphs via Transformer
This work addresses anomaly detection for applications in social networks, e-commerce, and cybersecurity, presenting a novel method for a known bottleneck.
The paper tackled anomaly detection in dynamic graphs by addressing challenges in node encoding and learning from coupled spatial-temporal patterns, resulting in a Transformer-based framework (TADDY) that outperformed state-of-the-art methods by a large margin on six real-world datasets.
Detecting anomalies for dynamic graphs has drawn increasing attention due to their wide applications in social networks, e-commerce, and cybersecurity. Recent deep learning-based approaches have shown promising results over shallow methods. However, they fail to address two core challenges of anomaly detection in dynamic graphs: the lack of informative encoding for unattributed nodes and the difficulty of learning discriminate knowledge from coupled spatial-temporal dynamic graphs. To overcome these challenges, in this paper, we present a novel Transformer-based Anomaly Detection framework for DYnamic graphs (TADDY). Our framework constructs a comprehensive node encoding strategy to better represent each node's structural and temporal roles in an evolving graphs stream. Meanwhile, TADDY captures informative representation from dynamic graphs with coupled spatial-temporal patterns via a dynamic graph transformer model. The extensive experimental results demonstrate that our proposed TADDY framework outperforms the state-of-the-art methods by a large margin on six real-world datasets.