Newsletters




Thwarting Fraud through Data Sharing


When people think of fraud, stolen credit card credentials come to mind, but fraudsters have developed new and innovative approaches to monetizing data about companies and individuals. A common and hard to detect fraud approach is create a fictional small business whose legitimacy may hard to verify for service and product providers.

XOR Data Exchange has formulated its Small Business Risk Exchange to help service providers determine both legitimacy and ability to pay for services, despite the challenges of little or no recorded credit history or employees.

According to Greg Bonin, chief operating officer, Exchange Services, XOR Data Exchange, preventing small business fraud can be particularly hard to do in the B2B space.  In a session titled, “Preventing Millions of Dollars in Fraud Losses Through Data Sharing” at Data Summit 2017, Bonin highlighted this use case and the approach that XOR takes detect fraud.

The legitimacy of small businesses can be extremely hard to verify, yet regulatory requirements require some verification due to KYC and AML requirements.  Often, there is no standardized spelling of a business names; many different businesses can have the same name; and the same business can exists in more than one location, among other challenges, said Bonin. In addition, he noted, the differentiation between fraud and credit losses not as sophisticated as it is for consumers.

While companies share data about wireless consumers to prevent fraud, companies in the B2B space do not, he noted.  XOR has developed a data exchange to do this. Bonin walked attendees through XOR's approach which includes the use of Lucene, a text matching engine, and data from a range of sources including demographic information, information from the request, publicly available internet data, and rating risky words versus less risky indicators.

Many conference presentations have been made available by speakers at www.dbta.com/datasummit/2017/presentations.aspx.   


Sponsors