From DPI Data to Warning: Real-Time Risk Assessment in the Fight Against Internet Fraud

Internet fraud represents a pervasive and escalating threat, undermining the security of individuals and organizations alike. To address this challenge, we introduce DiTing, a real-time intelligent risk assessment and early warning system tailored to protect potential victims. This system, termed Internet Temporal Behavior Analysis for Fraud Victim Risk Identification (IT-BrAVI), uses advanced analytics to scrutinize user behavior over time and evaluate the susceptibility of users to fraudulent schemes. Harnessing Deep Packet Inspection (DPI) data within an “operator-subject-time” analytical framework, DiTing adeptly identifies and flags risky URLs, performs nuanced internet temporal behavior analysis, and calculates real-time risk indices. Its implementation in Zhejiang Province has significantly curtailed the incidence of fraud, effectively safeguarding the financial assets of the populace. This paper discusses the theoretical foundations of DiTing, its architecture, methodologies, and the practical implications of its real-world application.