Using text analytics to more accurately detect fraud

Fraud detection using text analytics is gaining traction. We recently announced our initial focus on fraud detection through our partnership with IxReveal. Their advanced analytics software uses text to better detect fraud. In working together, we have seen a significant improvement in fraud detection accuracy when leveraging integrated structured data and text.

I expect 2009 to be a breakout year for text analytics. We’ve heard a lot about text analysis in the context of “voice of the customer” and I see a growing emphasis on eDiscovery. However, there are many other areas where text analytics will deliver considerable value. Fraud detection and quality reviews are two areas that should gain momentum next year. A good example of this momentum can be seen in our announcement today. We have entered into an agreement with MJM, Inc., the global leader in insurance fraud mitigation and claims investigative services, to deliver fraud detection solutions using text analytics technology for the insurance industry (Press release). The demand for new solutions to this growing problem is evident in their customer base. We expect to deliver rapid ROI in the form of reduced fraudulent payout, lower administrative costs, and timeliness of claim handling by leveraging text to more accurately identify fraud earlier in the claim process.

Having spent most of my CIO career managing business intelligence initiatives, I am excited about two key elements of this solution: the ability to easily gather and prepare data from various sources, and the ability to quickly get the text analytics engine up and running. Both of these solution strengths accelerate time to value - and in this market, time to value is critical.