The detection of cases of money laundering tends to become more accurate and agile based on new studies and technological launches. With a completely innovative approach, Computer scientists developed new software in England against this type of crime. The solution manages to be three times more effective than currently used methods.
The new feature allows the processing and analysis of a greater volume of data, which guarantees efficiency. Researchers from the IT Department at King’s College London created the software based on algorithms that quickly identify the “smurfing” or “structuring” technique. Like this, the system can monitor the actions of criminals by dividing large amounts of money into multiple smaller transactions and between different bank accounts.
Agility is the main differentiator of creation. The solution has the ability to analyze 50 millions of transactions in less than a second. This is essential to combat data on crimes involving money today: it is estimated that up to 5% of global GDP is laundered annually, according to data from the United Nations on Drugs and Crime (UNODC).
Software in practice
Until then, money laundering detection methods were based on rules and machine learning. This allows a type of automated alarm to be triggered in suspicious situations.. Some examples that may alert the system are depositing excessive amounts or processing large transaction histories.
Nonetheless, these conventional methods may be ineffective or slow for technologies used by criminals. Like this, the new software works using a large volume of data collected from different bank accounts and representing a complex graphic. This graph is the focus of the solution, allowing you to detect suspicious activity between accounts quickly.
Example: if a deposit of one million pounds is made, the system can monitor the exact destination of this transfer. In this way, the solution detects other correlated transactions even in cases where the money was divided between different accounts and expenses.
Available for free
New anti-money laundering tool software is open source and available for free. The approach carried out its tests using real data from an anonymous Czech banking institution, but, also, based on fictitious cases with predictions of common patterns and activities in the histories of this type of crime.
Details of this new feature had already been covered in an article published in the proceedings of the SIAM International Conference on Data Mining. The main researchers behind the innovative tool are Huiping Chen and Grigorios Loukides, together with Robert Gwadera and Solon Pissis.
