Big Data for Fraud Management

We will be conducting research to look into how big data can be used for better fraud management. We define big data as data of Volume, Velocity and Variety. Our premise is that more and more granular data from more sources allows banks, insurers, government agencies, e-Retailers to cut fraud losses more aggressively.We are interested in your thoughts around this topic.


Fighting the epidemic of prescription fraud with Big Data

As pharmacy benefit manager for 1 in 3 Americans, Express Scripts has visibility into 1.4 billion prescriptions annually. Our clients rely on us to help keep their pharmacy networks safe, and identify, investigate and deter potential fraud and abuse. Our program integrity team has dedicated specialists including nurses, law enforcement officers, auditors, accountants and pharmacists, that utilize proprietary, industry-leading data analytics to leverage insights from huge amounts of data in the fight against the national epidemic of prescription fraud and abuse. As you probably know, between 3-10% of all healthcare costs are lost to fraud and prescription drug abuse claims more lives annually than cocaine and heroin combined.
The team uses proactive data analysis to uncover patterns of fraud by identifying behavioral red flags. They receive thousands of tips and investigate hundreds of cases each year. Using the massive amounts of data gathered over the years, they have identified 290 potential indicators of pharmacy fraud alone. Geomapping, medical diagnoses, information on types and dosage of narcotics, and many other tools are utilized to identify potential patient fraud and abuse.
The information obtained is then further leveraged to develop predictive models that could help predict where potential future fraud hotspots. To learn more about the Express Scripts fraud, waste and abuse program, visit:

Really interesting and timely

Really interesting and timely topic. Rocketfuel's recent IPO demonstrates that using computer machine learning can unlock big data which they do for advertising. Is there more info on this research?