BIG DATA ANALYTICS FOR FINANCIAL FRAUDS DETECTION

Criminals and criminal organizations often make use of companies and other corporate entities to hide their identity, conceal illicit flows of money, launder funds, finance terrorist organizations, evade taxes, create and hide slash funds, commit bribery, corruption, accounting frauds and other financial crimes. These legal entities are frequently organized into complex ownership schemes set up in different countries, and with a “Chinese boxes” structure, in order to make it harder to determine who ultimately controls them and benefits of the illegal conduct. Currently there are a lot of competitors in the market of Financial Fraud Detection but the software that they propose are mainly oriented to supervise and manage the institutions’ internal compliance processes such as the management and transmission of Suspicious Activity Reports (SAR) instead of providing intelligence tools for proactively discovery potential threats and identify the final beneficiaries of illegal operations. Consequently there is a potential for Financial Fraud Detection focused on the on-line, real-time statistical analysis of transactions, operators behaviour, price movements and the use of data mining algorithms that work on heterogeneous sources of big data. After having described the schemes used for executing the three most relevant financial frauds this research proposes a novel approach for the detection of illicit behaviours and suspect transactions. The approach benefits of a multidisciplinary approach for the analysis of the big data streams coming heterogeneous sources such as TV stream, social media and public (official and unofficial) data bases