Combating Financial Fraud with Big Data and Hadoop

While you race around checking off items from your holiday lists, banks are just as busy with their fraud prevention efforts. According to a report by the Association of Certified Fraud Examiners, the typical organization loses 5% of its revenues to fraud each year, which translates to a projected annual fraud loss of over $3.5 trillion. Banks and other financial services companies are particularly vulnerable, due to the massive amount of financial data generated every day. Another challenge for this industry is the fact that the financial threat landscape has dramatically changed over the past few years, as sophisticated banking Trojans and new mobile threats pop up on a regular basis. The weeks prior to the Christmas holiday are typically a period of high malicious activity, and a new sophisticated banking Trojan is already gaining steam as the holidays approach.

The financial services industry is particularly complex, consisting of banks, credit card lenders, risk managers, and investment managers, and each of these sectors tend to approach fraud detection differently. The Consumer Sentinel Network, U.S. Department of Justice estimates that the total amount of credit card fraud alone worldwide is $5.55 billion.

As customer financial transactions occur at a much faster rate and are more data-intensive than ever before, banks and financial services firms are turning to big data, data scientists, and Hadoop to develop more sophisticated ways to prevent fraud. Companies who embrace commercial Hadoop distributions such as MapR are now able to:

Detect fraud more accurately. Have you ever traveled to multiple states in one day, only to find that your credit card has been declined? This is known as a “false positive.” Hadoop can be used to enable “behavioral authentication” which gives banks the ability to develop a more holistic view of their own customer behaviors and interactions. Data is analyzed from a variety of sources, from mobile data to social media activity. For example, a bank could learn from your Facebook activity and Foursquare check-ins (given the proper data privacy opt-ins) that you’re traveling, so your credit card activity would not be flagged as suspicious. In addition, banks can use Hadoop to find out how often a customer typically accesses their account from a mobile phone or PC – another method of gaining a deeper understanding of their customers’ behavior.

Identify fraud sooner. By using Hadoop to analyze vast amounts of data, banks are able to quickly identify fraud activity, and companies can be alerted to such activity in real-time. Traditional analytical approaches have often failed to accurately detect fraud as more exotic schemes are developed. This is because in many cases they were only looking for anomalies such as unusual login times or bad IP addresses, and they weren’t developed to handle large amounts of data and interaction patterns that change rapidly. With Hadoop, banking institutions can also review data across a number of different banking platforms, instead of analyzing each platform separately.
Predict future attacks. By leveraging big data and using predictive methods, financial institutions can anticipate attacks and work to proactively prevent them.

Utah-based Zions Bank, a subsidiary of Zions Bancorporation that operates more than 500 offices and 600 ATMs in 10 Western U.S. states, is one of the leaders in using Hadoop to help detect fraud. The bank uses MapR as a critical part of their security architecture. By using MapR, the bank is able to predict phishing behavior and payments fraud in real-time and minimize their impact, as well as run more detailed analytics and forensics. With MapR, Zions Bank is been able to significantly lower storage and capacity planning costs, as well as increase the speed of their analytics activities.

Detecting and preventing fraud is one of the biggest challenges that the financial services industry faces today. In order to stay ahead of the ever-changing landscape, banks are evolving their data architecture with innovations like Hadoop which allow them to keep up with the fraudsters to help prevent financial loss, increase customer loyalty, and gain greater insights that impact their business.
no

Streaming Data Architecture:

New Designs Using Apache Kafka and MapR Streams

 

 

 

Download for free