A new algorithm for money laundering detection based on structural similarity

Money Laundering (ML) is the process of cleaning “dirty” money, thereby making the source of funds no longer identifiable. Detecting money laundering activities is a challenging task due to huge volumes of financial transactions being made in a global market on a daily basis. This paper proposes a novel approach for detecting money laundering transactions among large volumes of financial data in an efficient and accurate manner. We propose a framework that applies case reduction methods to progressively reduce the input data set to a significantly smaller size. The framework then scans the reduced data to find pairs of transactions with common attributes and behaviours that are potentially involved in ML activities. It then applies a clustering method to detect potential ML groups. We present preliminary experimental results that demonstrate the effectiveness of the proposed framework.