Money Laundering Detection using Data Mining

Abstract: Money has given birth to numerous types of crime, one of which is Money Laundering. Money laundering is the method by which criminals disguise the illegal origins of their wealth and protect their asset bases, so as to avoid the suspicion of law enforcement agencies and prevent leaving a trail of incriminating evidence. To deal with this issue, the Reserve Bank of India (RBI), has introduced various guidelines to identify any sort of suspicious transactions. Once identified, they are forwarded to Financial Intelligence Unit (FIU). FIU studies the transaction and verifies its authenticity. However, this process is long, time consuming and not suitable to identify a certain type of transactions that occur in the system. To overcome this problems we propose an efficient Anti Money Laundering technique which is a composition of data mining approach and rule based engine. Data mining approach makes use of frequent association mining and graph analysis whereas Rule based engine works on the principle of rules getting fired whenever a suspicious transaction comes into main memory.

Keywords: Anti-money laundering, online payment, multi-agent, data mining, rule base agent.