A Probabilistic Model Using Graph Based Sequential Pattern Mining Algorithm For Money Laundering Identification

Money laundering an activity, which hides the source and origin of money in any banking or financial account of a country.  The countries financial stability or financial growth depends on the overall amount in the banks and finance organizations. The money holds the financial stability of any country and leads to changes the countries money value in international market. In past decades the criminals started hiding the source of income and the source of money from where the amount transferred to the finance account, which is illegal towards the financial rule of any country. This causes the threat to the financial stability of the country, because those amounts can be washed at any point of time to some other country. Also terrorist and criminals makes this kind of transactions to finance their clients to encourage terrorism. We focus on identifying the behavior of transactions and account holder in managing their accounts. There has been various methodologies proposed using data mining techniques, but suffers to identify the origin of money. We propose a graph based sequential pattern mining technique and probability model to identify the account from where the transaction originated. We generate a graph with many numbers of nodes and vertices for each account using the transactional data set, and using the account graph we generate sequential patterns and transition paths. Using sequential patterns and transition paths we compute the probability model to identify the origin of amount.