Sometimes, fraud is obvious.
Dennis Parker, insurance industry marketing manager at SPSS, a Chicago-based predictive analytics provider, remembers back to 1992 when he first started in the insurance business in the special investigations unit (SIU) of a large, nonstandard carrier. He describes seeing instances of what he calls "no duh" fraud, in which the indicators were so obvious that even the least-experienced claims professional or SIU team member should have been able to identify the claim as suspect -- such as when two obviously related individuals were involved in an auto accident with one another. 数据挖掘研究院
And while Parker says his team would identify such claims as potentially fraudulent most of the time, he still points out the cold, hard truth of the matter: "There's only one reason those claims were being perpetrated," Parker explains. "Insurance companies were paying them."
The industry has come a long way since then. Just about every carrier has a special investigations unit dedicated to fraud detection, and many have basic claims-scoring algorithms in place that automatically flag suspect claims. Some are able to access real-time databases to examine multiple claims filed by an individual. And although these capabilities sometimes are limited to specific lines of business, perpetrators of fraud no longer should expect insurance companies to pay out claims with no-duh indicators of wrongdoing. >>
But as insurers have improved their methods of mitigating fraud, perpetrators have improved their methods for committing it. "What we're starting to see is that those who are perpetrating fraud are becoming much more intelligent in the way they do so," says Michael Costonis, the Philadelphia-based executive director of Accenture's global claims practice. "They understand the scoring algorithms. They know what information they need to present at first notification of loss and then what to present at the next discussion to continually fly under the radar."
Fraudsters have indeed become adept at learning from experience and have leveraged that knowledge to increase their chances for future paydays. The insurance industry, however, has been slower to leverage its own institutional knowledge to keep pace with those who are walking away with claims dollars to which they are not entitled. According to Accenture's 2007 global claims study, only 17 percent of insurers currently utilize advanced IT tools to detect fraudulent claims.
Just as perpetrators of fraud today have learned from insurers, advanced fraud mitigation IT tools can help insurers learn from the perps. Some new solutions -- those with predictive modeling and text analytics functionalities, in particular -- allow insurers to mine data contained within their existing and incoming claims files to proactively, rather than reactively, establish patterns and associations, allowing carriers to develop better and quicker ways to accurately detect potentially fraudulent claims. 数据挖掘研究院
A New World of Fraud 数据挖掘研究院
Traditionally, according to Costonis, insurers have used an organizational approach to detect fraud. Potentially fraudulent claims are identified by scoring systems and red flags, which sometimes rely on manual processes and the expertise of experienced claims adjusters. Flagged files or those that tallied a certain fraud score are then sent to the SIU for further investigation.
But, "The new world of fraud is quite different," Costonis relates. "It's requiring much more advanced analytical and data-driven techniques to be able to get after these patterns, which aren't quite as obvious as they used to be." 数据挖掘研究院
Costonis estimates that less than 5 percent of insurers have data-patterning capabilities or the ability to find relationships between claims via data and text analytics. But that's not because insurers wouldn't welcome those capabilities. Unfortunately, many carriers would need to improve their paper-based claims situations in order to unlock the data found within their static, paper documents, Costonis explains, though advanced optical character recognition (OCR) technologies could help remedy the situation.
"The biggest challenge is just that [the necessary data] is not accessible to them in their systems," Costonis asserts. "As much as they'd like to have it, they can't get it because it's all in paper files." 数据挖掘实验室
Many companies that have already reduced their paper-based claims processes, however, have become de facto early adopters of data-mining technologies, Costonis notes. "Companies that are adopting the new core claims technologies -- the new types of platforms to remit their process -- are starting to look at [data mining functionality] as an add-on capability to help exploit the value in that data," he says. "It's the capability of the chosen few as opposed to the many."
Though not there yet, Los Angeles-based Farmers Insurance, a subsidiary of Zurich Financial Services ($67.2 billion in revenue), is moving rapidly toward adding text-mining capabilities and predictive modeling to the front end of its claims and fraud-fighting platforms, says Doug Ashbridge, director of special investigations at Farmers. Text analytics solutions extract factual information from unstructured text found in documents such as police reports, medical records and adjuster notes to establish patterns and identify trends.
The Farmers claims system has hundreds of structured data points, but Ashbridge still believes that a text analytics solution's ability to mine unstructured text could be beneficial to the organization's fraud-mitigation efforts. "By mining the text, you come up with more data points that were missed," he explains. "It also allows you to look for scenarios that are verbalized, such as the way in which an accident took place. You can't grab something like that out of data points."
Michelle de Haaff, vice president of marketing and products at Attensity (Palo Alto, Calif.), a provider of text analytics solutions, including one specifically designed for workers' compensation claims, says that most insurers do not have processes in place to look at the unstructured data that represent a large portion of the total data housed within a claims file. "In general, an insurer may have excellent processes in place regarding fraud, but they're missing a big, important piece that makes a huge difference in their ability to identify fraud on a mass level," she contends. 数据挖掘研究院
While Farmers may soon implement data-mining technologies such as text analytics, it already has an active and effective fraud-mitigation operation, according to the carrier's Ashbridge. In late July, for instance, the carrier won a fraud lawsuit against Hollywood Auto Collision. The case was the first tried under a 1993 California statute designed to augment law enforcement efforts to prosecute defrauders, according to a Farmers press release.
Zero-Tolerance Policy
Farmers enforces a zero-tolerance policy regarding fraud, according to Ashbridge, who has been associated with the company's SIU for 25 years. While acknowledging the organization's aggressive approach to fraud investigation and the diligence of its well-trained SIU and claims adjuster workforce, Ashbridge credits technology with helping to improve fraud mitigation over the years.
"We have a very robust back-end process where we're able to pull in large amounts of data from our databases, massage that data and show trends and patterns -- areas where we should start to look for fraud," Ashbridge says. Currently, he relates, the company uses ISO's (Jersey City, N.J.) link analysis technology NetMap and i2 investigative analysis software to quickly identify trends and patterns within its own past claims data and information drawn from outside databases, such as the consolidated All-Claims database cooperatively maintained by ISO and the National Insurance Crime Bureau.
McLean, Va.-based i2, which is owned by ChoicePoint (Alpharetta, Ga.), allows Farmers to graphically visualize claims patterns, Ashbridge adds. "NetMap pulls all the information together and i2 allows you to display it in a way that makes sense," he says. 数据挖掘研究院
Accuracy First, Detection Second
But making sense of the data is harder than it sounds. Analyzing data to detect trends and patterns -- as opposed to simply using old fraudulent claims to directly detect new fraud cases -- is only now an emerging industry best practice. "There is a certain degree of fallacy in using your past fraudulent claims to predict future claims," Accenture's Costonis says. "The key thing you should be trying to recognize is a pattern and how it evolves over time as opposed to trying to predict the specific fraud indicators." 数据挖掘研究院
Columbus, Ohio-based Nationwide ($22.3 billion in 2006 consolidated total revenue) employs a similar philosophy and utilizes several different technologies to defend against fraud. The insurer uses ISO NetMap for links analysis capabilities and, in 2005, implemented a solution from Chicago-based Magnify (another ChoicePoint company) for predictive modeling, according to Donnie Kearns, SIU director, Nationwide.
Kearns says Nationwide leverages predictive modeling, in part, to find trends in past fraudulent claims. "Leveraging predictive modeling improves the accuracy of fraud detection. In short, it narrows our focus on fewer claims," he explains. "Claims professionals can't really appreciate patterns because they are handling so many claims, and each is treated as an individual transaction. They don't get a chance to step away from the forest to see the trees, so to speak." 数据挖掘研究院
Kearns and others caution, however, that using only predictive analytics to detect fraud could be ineffective. "As powerful as modeling is, you shouldn't let that work alone," he stresses, adding that fraud mitigation is more than just learning from past mistakes, especially when it comes to predictive modeling, which uses past evidence to predict future results. 数据挖掘研究院
"Predictive modeling is primarily going to tell you what you know or have experienced," Kearns continues. "You shouldn't stop there but, rather, employ other tactics. Those efforts should make your model stronger over time." 数据挖掘研究院
And few observers doubt that current fraud predictive models have room for improvement. According to SPSS' Parker, many experts within the insurance industry believe that 12 percent to 15 percent of all claims should be referred to the SIU as potentially fraudulent. However, only a small percentage of those are ever actually identified as such.
"That means that when you go to create a predictive model based on that half of 1 percent, you're not including all those [fraudulent] claims that you failed to identify," Parker explains. As far as the predictive model is concerned, those unidentified claims could be considered valid, throwing off the accuracy of the system.
Even the SPSS solution suite itself, known as PredictiveClaims, does not rely entirely on predictive modeling, Parker notes. He says the solution first scores claims based on a business rules engine and then "optimizes" those scores with a predictive model. 数据挖掘研究院
A Human Touch 数据挖掘研究院
Still, no single technology currently can act as a complete fraud-mitigation solution. Therefore, attacking fraud multilaterally is the best approach, according to most experts. "You shouldn't rely on a single approach to get files into the SIU," Nationwide's Kearns emphasizes. "We want to make sure we do as much due diligence as possible prior to investigating claims." 数据挖掘实验室
Despite all the new technological advances out there, human intervention remains a key component of any organization's fraud detection and prevention efforts. "While the predictive modeling does the heavy lifting, we still rely on human intervention to validate the indicators and determine whether a claim warrants further investigation," Kearns relates. 数据挖掘实验室
Many times, organizational structures and strategies can determine the success of an insurer's fraud-mitigation efforts more than any set of technology solutions could. "The technology is never going to replace the human decision-making factor," Accenture's Costonis contends. "The real value of the technology is being able to surface the pattern and the changes in that pattern over time. You'll still need an experienced person to interpret that pattern and decide what to do with it."
Nationwide relies on human discernment when making the final determination of whether a claim warrants further investigation. "We understand that technology is not perfect and rely on human intervention to determine whether indicators are present and valid," the carrier's Kearns says.
Insurers would recoil if a software vendor decided which claims were and were not fraudulent, adds SPSS' Parker. Further, almost all insurers have different philosophies and mind-sets on what claims should be referred. "What a vendor can do is show insurers which claims are potentially fraudulent, show them the reasons why and then let the investigation run its course," Parker says.
Many experts agree that facilitating cooperation between SIU staff and claims professionals is a key fraud prevention best practice for insurers. Just as the "throw it over the wall" approach to more general IT-business relationships is ineffective, so too is the once-and-done handoff approach to claims fraud investigations.
According to Accenture's Costonis, more-efficient fraud detection operations facilitate collaboration between claims professionals and the SIU, with each group consistently providing its expertise. Otherwise, "a lot of opportunity gets lost in translation -- in the handoff between the claims professional and the SIU department," he says. 数据挖掘研究院
At Farmers, SlU staff randomly review claims files that were not referred to the department, looking to see if signs of potential fraud were missed by claims adjusters. "That's done in order to spot areas where people are not identifying the indicators, and it tends to be targeted toward areas where we are seeing a higher-than-normal occurrence of a certain type of fraud," explains the carrier's Ashbridge. "It gives claims professionals proactive assistance in identifying things from a human standpoint." 数据挖掘研究院
Ashbridge describes the carrier's SIU as an adjunct to its claims department, with investigators and adjusters working in tandem. In fact, investigators don't take over claims files referred to them. Instead, adjusters maintain that management responsibility. "The SIU assists and guides the claims professionals through the investigative process," Ashbridge relates. "By maintaining those [potentially fraudulent claims] files, it keeps them keen and in the game. It keeps them active in the fraud aspects of handling claims." 数据挖掘研究院
