Fraud Detection Through Large-Scale Graph Clustering with Heterogeneous Link Transformation

This article likely presents a novel approach to fraud detection by leveraging graph clustering techniques. The use of heterogeneous link transformation suggests the method can handle diverse data types and relationships within the fraud network. The focus on large-scale graphs indicates the method's scalability and potential for real-world applications.