sharing and analyzing data to reduce insurance fraud

insurance fraud is a multi-billion-dollar problem.  fraudulent practices occur frequently and often repeatedly. fraud can be detected and prevented if appropriate data is collected, analyzed and shared among insurance companies.  appropriate decision support and analytics can be developed to routinize fraud detection. creating these decision support capabilities involves addressing managerial, technological, and data ownership issues. this article examines these issues in the context of using new data sources and predictive analytics to both reduce insurance fraud and improve customer service. evidence suggests that appropriate sharing of proprietary company data among industry participants and combining that data with external data, including social media and credit history data, can provide advanced data-driven decision support. cooperative development and deployment of predictive analytics and decision support should reduce insurance costs while improving claims service. a process model is developed to encourage discussion and innovation in fraud detection and reduction.