The term "social networking" is used quiet freely today to represent a myriad of functions. Most commonly, social networking is used in context to social interaction - both online and off. The best and most used example today for a social network is Facebook.
However, the term social networking has also been used traditionally to describe a method for analyzing social networks that exist outside of Facebook and other online channels. In particular, investigators use social network analysis to identify entities working conclusively together to perpetrate fraud. This usage is somewhat of a misnomer, since what is really being utilized is a statistical method called link analysis.
Wikipedia defines link analysis as a technique used to evaluate relationships (connections) between nodes identified in terms of network theory. Relationships may be identified among various types of nodes (objects), including organizations, people and transactions. Link analysis has been used for investigation of criminal activity (fraud detection, counterterrorism, and intelligence), computer security analysis, search engine optimization, market research and medical research.
The primary reason this is important is that is many cases social networking is presented as a method used to identify aberrant or fraudulent behavior. Where in actuality it is only providing a method of viewing associations (or links) of an entity that has already been identified by other means.
This can provide value to investigators for case work, but really does not help identify more cases or collusion. However, if actual statistical methods are utilized to identify patterns of behavior, social networking or link analysis can provide tremendous insight on collusive behavior.
For example, it can show you how providers are billing the same member/beneficiary for the same procedure. Or, perhaps, analyze provider referrals to durable medical equipment suppliers. This makes it possible for link analysis to really provide some true target identification for collusive fraudulent activity.
To wrap it up: make sure, if considering using social networking for fraud detection, that the analysis is actually detecting fraud and not simply helping investigate it.
