Using Social Media Network Analysis to Detect Bank Application Fraud

Social media.jpg


Vipul Srivastava
Fraud Analytics Professional

As of April 2018, the global population is 7.6 billion and 4.2 billion of us reportedly use the internet. There are 3.3 billion active social media users, which means that 70% of people using the internet also have one or more social media accounts. This number is expected to increase to 90% in the next three years. As a fraud analytics professional, this tells me the majority of people who have access to the internet may also be trackable on social media. This presents an exciting opportunity to detect and minimize bank application fraud.

Banks and financial organizations are increasingly exposed to bank application fraud risks when fraudsters penetrate the system by submitting false information and fake documents. Once they are accepted as customers and provided credit lines, they can defraud a bank and cause significant losses. To counter bank application fraud, many banks deploy application screening systems which trigger alerts when an anomaly or pattern is identified. These alerts are then actioned by operations in order to approve or reject an application.

Companies are increasingly using social media to check the profiles of bank applicants, however, banks only check the given profile information of an applicant. For example, operation teams are checking Facebook or other social media platforms to do a very preliminary level screening- like matching the photos of applicants or checking basic details like living statuses and work profiles.

Is this good enough? What if a fraudster creates a fake profile just like he created on in his application? A check on his fake profile will only validate the details mentioned in his application. Thus, institutions may view this application as genuine and could fall victim to the trap.

A new and robust way to assist operations in identifying fraudulent bank applications is to calculate the personality traits of an applicant based on his social network. It involves not only examining the social media profile(s) of an applicant but also screening his network of friends and connections. It's often said that people are reflections of their friends, and this process gives a better picture of an applicant. It’s easier to make a fake profile, but it is tougher to add genuine friends.

For example, if an applicant claims a certain lifestyle or social background, but most of his friends are of a different nationality, come from different social backgrounds and live different lifestyles, then this could be a red flag. Similarly, frequent abusive and derogatory posts or tweets from his network of friends may be an indicator of one’s personality.

Calculating the personality score of an applicant based on his network profile or activity can provide useful insight on whether to onboard a new customer or not. Companies should invest in deploying smart solutions to perform this analysis in order to mitigate bank application fraud risk.

Vipul Srivastava works in the fraud risk department of a global bank and has keen interest in exploring new ways to mitigate fraud risk.