Printer Friendly

Order from chaos: can semantic technology solve the unstructured data riddle?


In the early 1990s, the claim industry received a startling wake-up call. There were thousands of accidents involving SUVs that flipped. While the industry was aware of the issue, claims and information as to the scope of the problem came in slowly over a period of several years.

Additionally, the information needed to assess the financial impact of the situation was often buried in adjuster notes. As such, finding it required time and a considerable amount of manual labor. That meant insurers had a difficult time identifying the magnitude of the potential cost. Soon, that handful of rollover claims grew to become one of the most significant collections of product failure claims to hit the insurance industry. Insurers recognized that there had to be a better way to get a handle on these types of situations. They realized that new technologies, procedures, and approaches would need to be developed and implemented.

Nearly 15 years later, that new approach is here. The same technology that makes it possible for thousands of young people to share information on social networking sites also helps claim executives spot the patterns in their claim files that, when left undetected, could send loss expenses spiraling.

Known as semantic technology, the latest evolution of web interaction is helping forward-thinking insurers to identify better ways to integrate data; identify patterns and emerging issues such as exploder claims; improve strategic decision making; create greater efficiencies; and lower overall claim costs.

This type of technology evolved as a solution to a problem. The problem was how to intelligently understand the content on the Web in order to allow users to interact and communicate with that content as well as with one another. Semantic technology describes the meaning of the data and how it can be used at execution time, separately from the data and the application code. This allows people, machines, and applications--with no prior knowledge of each other--to understand and reason with one another at execution time. It allows a real-time declarative approach to adding new data, building new applications, and in combining existing applications into new products.

Semantic technology is also a key component of Web 2.0 (and beyond), the newest iteration of the World Wide Web typified by Google-like intelligent search engines, social networking sites such as Facebook and MySpace, and blogs. Web 2.0 is best described as the interactive content on the Web, where the users are partially or largely responsible for what the web site is presenting. Whenever people use a search engine, type blog entries, or update their Facebook pages, they are using Web 2.0 tools powered by semantic technology.

Tip of the Iceberg

The genesis for the use of these technologies in the insurance industry is the growing recognition that insurers only use a small set of available information when managing claims. This information, known as structured data, is typically in the form of standardized fields in claim systems. While useful, such data only represents a fraction--by some estimates, only a mere two percent--of information that is available in their claim files. In other words, claim management and other interested parties are only seeing the tip of the iceberg. The remaining 98 percent--known as unstructured data--is found in adjuster notes, e-mails, imaged documents, case-manager notes, recorded statements or digital audio files, and digital photographs. Because it cannot be "read" by conventional software systems, it must be analyzed manually.

In addition, available data is typically accessible only to a select set of users. Often, departments within the same company remain highly compartmentalized, so the claim department may not always know what is happening in underwriting, legal, accounting, or risk management. These are all areas that have considerable relevant and useful information to contribute to the claim adjudication process. Semantic technology can transform this compartmentalized data into data "legos," allowing end-users to reassemble data dynamically into new and meaningful categories of information.

While understanding the communication and entertainment value of Web 2.0, many insurers remain unsure as to what, if any, real-world application semantic technology holds for the industry. Consider the previously mentioned rollover example. What if the insurance company decision-makers had fast access to unstructured data trends from across all their claim office locations? What if the raw adjusters' notes were trans formed into nuggets of valuable information that could be easily searched and navigated? What if a colleague in another state could attach a virtual "sticky note" or a tag indicating that he had seen similar accidents and ask if more research was needed to determine if there were commonalities? The situation may have been identified, quantified, and mitigated much quicker.

With semantic technology, the awareness of useful information and communication among key parties is greatly increased. Connecting this information helps to highlight the business significance of events that occur in various parts of the organization but are related to the same claim.

The Cool Factor

Semantic technology uses a methodology that takes structured and unstructured data and runs it through a transformation process. In effect, the data from an array of sources, including adjuster notes, spreadsheets, PDFs, scanned notes, image files, correspondence, and existing data warehouses are combined into HTML or other W3C open standard formats for semantic processing. This gives every user access to their own data warehouse and makes the data scalable, searchable, more readily shared, and in general, much easier to use.

In short, this technology provides a Google-like function for claims, allowing the user to search and navigate through data from transactional systems, multiple lines of business, and multiple applications. This approach provides the adjuster with a 360-degree view of the entire claim profile. It also allows claim executives and decision-makers to research an entire claim inventory in just a few seconds, using any number of searchable terms and concepts. For example, professionals can search for all accidents involving an incidence of road rage; all of those with a certain make of automobile; or perhaps all involving specific attorneys.

This type of access means that trends and useful information for a number of departments can be more readily discerned. This information can then be used in underwriting to spot increased risk, to identify exploder claims for special handling, to detect patterns that indicate potential fraud, to find subrogation opportunities, and much more.

With this approach, the employees on the front line of claim processing have the ability to lead the company to a better approach to process claims and manage costs. In addition, managers and executives have enhanced access to information that can help them intervene and take appropriate actions to prevent or better prepare for catastrophic claims.

Virtual Sticky Notes

The latest generation of semantic technology provides the capability to capture and "tag" information of interest. Tagging represents the ongoing evolution within semantic technology. Think of tagging as affixing "sticky notes" or red flags for key data. With this technology, relevant findings can be saved as a category, book-marked, kept private until more information is gathered, or communicated to the rest of the organization.

Tags enable structured fields to be added to data dynamically. The value of this capability is that when trends or patterns are identified, appropriate action can be taken immediately. This would include subrogation referrals, special investigations, underwriting alerts, and so forth. Such an approach allows for greater efficiencies while creating distinct niches of expertise that can lead to improved settlement performance.

What's Next?

The integration of semantic technology in the insurance industry represents more than the adoption of a new software or system. It signifies a fundamental and lasting change in the way information is gathered, analyzed, and shared throughout the enterprise. As such, it must be approached as organizational change management. The focus is not on learning a new technology but on encouraging and fostering a lasting change in behavior. While representing a seismic shift in the way business is conducted, it is an important journey, and is one that must start now. Not only will it benefit existing employees, but as baby boomers retire, it will also provide a common language and familiar skill set for a new generation of workers. This new generation will be ready to use semantic technology immediately. The industry is already discussing Web 3.0, which may include artificial intelligence; automated reasoning and inferencing; access to semantically aware services and modular web applications; and advances in semantically aware user interfaces. In short, Web 3.0 will provide the kind of knowledge that can help industries transform and compete in an increasingly global marketplace. Those insurers that seek to take steps to meet the opportunities of today, as well as the coming changes will be well positioned for success today and in the future.

Tip of the Iceberg

Most MIS Reports Only Show Structured Data--2% of available information

* Data of Loss

* Claim #

* Accident State

Missing 98% go of Unstructured Data

* Adjuster Notes

* E-mails

* Imaged Documents

* Record Statements

Improved Results

* Reduced loss Cost

* Greater Insights

* Competitive Advantage

* Actionable Intelligence

A Road Map for Web 2.0

Because of semantic technology's ability to maximize the use of available information and share data across the enterprise, many industry analysts believe it will be the next major IT evolution within the insurance industry. The overarching question for insurers will be how to implement this innovation. Here are a few things to consider:.

* Analyze the company's needs. Are departments and personnel communicating as they should? Do bottlenecks exist?

* Consider the time it takes to review claims manually. What if that process could be accelerated by searching through thousands of claims in milliseconds?

* Talk to your IT staff, including data warehouse personnel. Ask for suggestions as to how existing technologies can blend with Web 2.0.

* Hold discussions with your rank and file. Find out what they need and want in a new approach.

* Understand how the new technology will affect employees. Change performance standards and training accordingly so that employees have assurances that using the new system will ultimately help them meet the goals and objectives of their managers.

* Develop protocols for working with semantic technology. Web 2.0 opens a world of opportunities for insurers. Organizations need systems to review what has been tagged, vet findings to ensure appropriateness, and provide guidelines and training for users.

* Determine whether you will build a system, buy one, or hire a consultant to develop the semantic program, Semantic technologies are also available as Service as a Software {SaaS) formats. This type of technology is cost-effective, works with existing legacy programs, and is relatively easy to implement.

Bill Nadal is senior vice president and chief technology officer at Full Capture Solutions, Inc., which develops and provides analytic technology and applications for the insurance industry. He may be reached at 415-892-1952,
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Feature Story
Author:Nadal, Bill
Date:Sep 1, 2008
Previous Article:Proactive investigations: curiosity kills the cat--and the comp claim.
Next Article:Sifting through the ash: misunderstanding building materials can leave you burned.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |