Washington fraud-busters turn to social analytics

Faced with a US$1.8 billion budget deficit plus staff cuts and pay freezes in the civil service, losing millions to fraud is the last thing the state of Washington needs. In an interview with FutureGov in Seattle last week, fraud detection agents from Washington's Department of Labor and Industries (L&I) revealed how social networking business analytics is curbing the rise of compensation cheats.

The problem is not new, although it is being exacerbated by the weak economy. L&I set up the Fraud Detection and Prevention Unit six years ago amid criticism that the issue was not being effectively tackled. "We think that we were only correctly identifying cases of fraud 50 per cent of the time. Often we were taking time out of an employee's schedule to investigate him or her, when there wasn't a problem. There was no ROI," said Hammersburg.

In the past, L&I used a variety of inhouse technologies when investigating an employee or firm; rules-based programmes were used so that when questions were answered in a certain way, an investigation would be triggered. Now the department is replacing the systems its auditors use with a single platform, developed by SAS, to cut down on the time taken to investigate a case and make audits more efficient and effective.

The new system uses social networking analytics to assess the likelihood of fraud, and to determine whether that case could be related to others that form part of a crime syndicate. "Fraud is decreasingly about a single individual or company. It's become a way of business that grows as others are recruited," said Doric Olson, Fraud Detection/Operations Manager, L&I. "Our new system can find connections between cheating individuals."

Read more.