Towards Unified Learning Framework for Graph Anomaly Detection. This project aims to build a unified framework to detect anomalies in networked data, such as fraud in financial systems or cyberattacks

Description

Towards Unified Learning Framework for Graph Anomaly Detection. This project aims to build a unified framework to detect anomalies in networked data, such as fraud in financial systems or cyberattacks in computer networks. Existing solutions heavily rely on domain-specific data and costly computations to establish scenario-specific models, significantly limiting their applicability and generalisability in data-scarce, privacy-sensitive, or rapidly evolving scenarios. This project expects to design novel techniques to build a unified framework that can generalise across different application domains, data types, and anomaly types. The framework should benefit domains like finance, cybersecurity, and environmental monitoring, enhancing security and efficiency for governments, businesses, and communities.. Scheme: Discovery Early Career Researcher Award. Field: 4605 - Data Management and Data Science. Lead: Dr Yixin Liu