The insurance industry remains under continuing pressure to increase operating efficiency through better data utilization. At the same time, insurers must also manage a wide range of risks. By using social network analysis (SNA) to root out fraud and criminal behavior, carriers can increase efficiency and manage risk simultaneously.

In recent years, a new brand of criminal has emerged that is immune to conventional risk scoring. Traditional data and record-matching techniques struggle with poor data quality, missing data and an inability to root out deliberate attempts by criminals to hide identities. Legacy systems mostly resort to inexact or fuzzy matching, which generates reams of false positives that few institutions have the staff to thoroughly investigate.

Social network analysis, also known as link analysis, is a powerful tool in understanding the structure of social and organizational networks that are often connected to criminal behavior. SNA maps and measures relationships and flows among people, groups, organizations, computers or other information/knowledge processing entities. The nodes in the network are the people and groups, while the links show relationships or flows between the nodes.

SNA provides both a visual and a mathematical analysis of complex human systems. This analytic approach has practical importance because SNA tools combine data extraction, manipulation, analytic and visualization tools to distill massive databases into a visual representation of unusual linkages. It tells us who knows whom, who calls whom and who does business with whom.

By monitoring the communication patterns between network nodes, its structure can be established. Identifying the structure of an insurgent network enables identification of critical nodes and their relationships.

The basic analysis needs a predictive analysis component to understand the "pattern of life" within the network. Together, network analysis and predictive analysis enable financial institutions to identify the network, determine critical targets, and predict when and where targets may take advantage of an opportunity.

Standard rules-based systems can't unearth "first-party fraud" and "bust-out fraud" where criminals establish accounts for the sole purpose of committing fraud.

A classic example is found within the credit card industry. TowerGroup, a Needham, Mass., research and analysis firm, projects that total card credit losses for issuers of U.S.-branded cards will peak at $55.6 billion in 2009. Rules-based systems are looking at more traditional types of risk, such as poor credit. With SNA, fraud-based risk can be seen by investigators, making it easier to uncover previously unknown relationships and conduct more effective investigations.

Hybrid Approach

Analyzing social relationships could be particularly useful in combating organized crime rings. SNA uncovers connections that better assist investigators and analysts in producing usable intelligence. This technique can expose fraud faster, identify indirect crime and deceptive patterns and leverage information linking fraudsters to illegal activities.

While rules-based analysis can detect some of this activity, it can't detect the most sophisticated crime. By integrating advanced analytics with existing business rules, end users can incorporate clustering analysis, mean and standard deviation, data mining and predictive analytics to create a powerful ally in fraud prediction and protection. Using this high-level, hybrid approach to network analysis, insurers can optimize their existing investment and evolve their detection process to incorporate more intelligence while refining the alert monitoring and detection process.

SNA methods in the context of an ongoing fraud or criminal investigation can eliminate antiquated guesswork and ad hoc reporting. While those methods can be economical, they don't provide the flexibility to follow a trail of links that may not be immediately apparent. An interactive reporting system enables investigators and analysts to query data and search for interesting or unusual connections.

Overall, SNA can reduce the time to detect fraudulent situations while automating the investigation's time-to-resolution as companies utilize the ability to pick up on subtle, illegal behaviors that typically went undetected.

A robust SNA solution needs two key components:

1. A platform that marries the three components of detection, alert management and case management, while providing category-specific workflow, content management and advanced analytics. These components are fully integrated with SNA, and offer both top-down and bottom-up functionality in making hidden and risky networks more visible.

2. Advanced, large-scale network analytics that work across internal and external data sources to link customers and accounts based on common attributes or more subtle patterns of behavior. By integrating SNA into the entire fraud framework process, previously undiscovered alerts and flags can be fed back into the alert monitoring process to fine-tune the ability to detect fraud.

While social network analysis has been present in some form or another for decades, leading-edge companies will make their mark in the coming years. Rules-based programs don't factor in people's relationships. SNA does. You can learn a lot by tracking and observing these interactions.

This article can also be found at