Breaking Through Opacity: A Context-Aware Data-Driven Conceptual Design for a Predictive Anti Money Laundering System

摘要

We studied the problem of money laundering. Current anti money laundering artificial systems solve this problem by tracking transactions. However, several inherent aspects impede a full uncovering of the underlying process; most arguably the abstract nature of transactions and the ‘legitimacy dilemma’. We circumvent these issues by proposing a context-aware and data-driven software/hardware system that tracks physical money rather than abstract transactions. The proposed system identifies banknotes by their unique serial numbers, allowing a meta data component that accumulates further contextual information to better detect unusual patterns of transactional behaviours. In addition, using ‘social network analysis’, the system can be used to predict future money laundering processes.