The IT giant took the wraps off a new IBM Loss Analysis and Warning application that uses data mining, scoring and anomaly discovery algorithms to help property and casualty insurers weed out fraudulent claims.
Insurance fraud is a growing problem, costing US P&C insurers tens of billions of dollars annually. According to the National Insurance Crime Bureau, 10% of all insurance claims are either inflated or outright fraudulent, resulting in around $30bn in payouts.
IBM Loss Analysis and Warning lets P&C insurers build up compliance models for particular claims, individuals and organizations and then matches, compares and ranks claims activity against this model to identify suspicious claims. For example, the system would automatically red-flag an auto repair shop charged for towing in all its claims while other shops has less towing claims.
The system can also be configured to immediately analyze a claim as it is filed as part of the claim review process, thereby helping insurers avoid the so-called pay and chase dilemma faced by insurers, where they find themselves investigating a potential fraudulent claim long after it has paid out.
There is also an automated investigator-in-a-box tool that automatically applies parametrized analysis modules and logic to the data to identify questionable claim behavior. IBM said this capability frees up time for investigators to focus on case development and pursuit.
The application was built by IBM’s Center for Optimization and meld’s the company’s analytic and data management technologies, industry domain expertise and mathematical modeling know-how into business-specific solutions.
The underlying technology is the same as IBM’s Fraud and Abuse Management System for the healthcare sector.
IBM Loss Analysis and Warning is available as a licensed or IBM-hosted (on-demand) application.
IBM Loss Analysis and Warning is part of IBM’s risk management optimization suite. IBM’s Center of Optimization also offers solutions in the areas of supply chain management optimization, marketing investment optimization and dynamic pricing optimization.
