Jun 5, 2016

Jun 5, 2016

Insurance Fraud As A Terrorist Activity - The Intelligence Point of View


As many of us already know, insurance fraud is defined as anyone who seeks to benefit from insurance through making false claims of loss or injury. This may occur when a claimant attempts to obtain some kind of benefit or advantage to which they are not otherwise entitled, or when an insurer knowingly denies some benefit that is due.

How common is insurance fraud? According to a recent research, conducted by AXA, we’re talking about one in three people (!) have committed insurance fraud. Needless to say, the true scale of this problem is not shown, as many people would never openly admit to breaking the law.

So, you’re probably asking yourselves; how can we combat insurance fraud? First of all: you’re not. You’re not going to combat insurance fraud, you’re going to treat it carefully. We’re talking about understanding the situation, seeing the bigger picture and realizing the scale of this phenomena. Then, you’ll have to start treating insurance fraud as if it was a terrorist activity -  specifically meaning that you are not fighting an army or specific country, you’re fighting guerrilla-style criminals, individuals and double agents. As an Israeli company, we’re surrounded by some of the best and most inspiring intelligence agencies/personnel. We allow ourselves to brag that there is no better country in the world in combating terrorist activity, and by combating we mean; going through the entire intelligence cycle - collection, processing, analysis, dissemination, and feedback. Just like a military operation, we’re using different types of intelligence, whether it be HUMINT (Human Intelligence), SIGINT (Signal Intelligence), OSINT (Open Source Intelligence), or everything together in order to gather the data that is further utilized by our analytical engine. In order to bring all pieces of data to a common denominator and make them usable by computer systems, we at GetMeIns use ubiquitous language (or ‘ontology’) in our insurance process.

Ontology has its roots in philosophy and deals with the question “What really exists?”. It defines entities, relationships between them, various concepts or ideas describing some aspects of specific domain (in our case, insurance). With the help of ontology and domain modelling techniques, we define the fundamental categories or classes into which various objects we encounter during the information retrieval process usually fall.
There is an ongoing process of knowledge crunching that leads to discovery of new concepts and, as a result, new entities and attributes that enrich our ontology.

Using ontology helps us make data fusion from multiple information sources to accomplish a puzzle. And when all the pieces of a puzzle are in place we can eventually predict the next fraud case! In future posts we’ll elaborate more on information discovery process and ontology analysis.  

Utilizing the intelligence approach allows us to gather the necessary data and analyze when there’s still time to act (as opposed to after the claim, which is what generally exists in today’s market).

Our next blog-post will focus more on the different types of insurance fraud and how to combat them. We’ll discuss some of our unique working methods and tools and explain our different approach.