DataGYM » Showcase » Case Studies » American Express

FACT SHEET
Richardson, Texas; August 16, 2000

CIANT Completes Fraud Appliance Pilot At American Express

Resources required for hands-on Analyst review reduced by 33%

How to stop premeditated fraud attack on a business?

Is it possible to reduce losses from:

• Issuing lines of credit to a known or potential fraud?
• Authorizing a transaction one will never collect on?
• Sending merchandise on credit to someone determined to rip off a business?

Customers can now use a number of channels simultaneously to touch a business. In the "Old Economy" business model, or in the B2B and B2C "New Economy" business models. Customers have only one thing in their minds, to quickly and efficiently complete their transaction.

Technology is geared to deliver this. There is a minimum set of information requested from a client to verify the authenticity of the caller and approve a sale or service. Anything else might inconvenience the clients or be uneconomical for an enterprise.

Fraud perpetrators know this very well and their objective is to find those weak links in this technology and decisioning architecture and take advantage of them.

Such calculated and premeditated attacks can be stopped by a Fraud Appliance based on CIANT DataGYM Instant Matching Engine.

The DataGYM Fraud Appliance attempts to provide such an environment to an organization. The DataGYM Fraud Appliance is specialized server with a database environment built for this purpose. The server can be hooked up into an Internet or Intra-net environment with firewalls and other protective layers. The server is designed to accept any number of data elements that are parsed and compared against a stored database. A probability is assigned to the match based on the information available. This information is returned to the requesting environment, which decides the course of action based on this result.

The same tool can be customized to define a centrally deployed solution where only the final decision is returned to the requesting system, web interface or Call Center.

There are three components in a DataGYM Fraud Appliance:

The Input Process - accepts data and standardizes it to make it easy for the Instant Matching Process to handle it. Create a number of Phonetic-Key(s) (based on a proprietary Phonetic Algorithm) and invokes the Instant Matching Process.

Facts:

1. Table based parsing engine to parse all critical data elements.
2. Table can handle English to Japanese Kanji data with equal ease.
3. Input data automatically validated to prevent garbage-in/garbage out.
4. Address Correction or Address Standardization may be incorporated.

The Instant Matching Process - extracts all information from the database that might potentially match the input data and pair the data based on a customized sequence. There are a number of matching algorithms based on probabilistic and lingual references. All these can be changed and customized based on country of application or on Client's requirements.

Facts:

1. Phonetic use of language to allow spelling distortions and variations.
2. Probabilistic algorithms to determine match probability.
3. Cut-off probability at match levels reduce false-positive matches dramatically.
4. First Match/All Match/Best Match options.
5. Unlimited number of match hierarchy and unlimited number of matching source tables.
6. Nick Name tables and X-Matches to make sure name or data components have not been switched.
7. Phonetic Tables can be changed for each market due to language rules.
8. Match data can be given different weights to emphasize particular words.

The Output Process - As soon as there is a match, the results are formatted to return to the calling environment via this process.

Facts:

1. Return first match with probability.
2. Return all matches by importance and probability.
3. Return best match by importance and probability.
4. Can be customized to return decision code to calling environment.

There are two other components that are required to establish and maintain the Fraud Appliance:

DataGYM Conversion

used to clean, parse and standardize data before loading it into a data repository. More details.

DataGYM House-holding

used to establish all the phonetic rules, matching hierarchy and matching data source details. More details.

This environment is a win-win for all since almost everyone doing business over the Internet, over a telephone or traditional Telesales organization can leverage this platform.

For small businesses this service is also available via the web and access is provided to a cooperative fraud database created from pooled fraud experience of various Clients.

Like the Fraud Appliance applications, the DataGYM platform can be used for instant decisioning purposes real-time in a variety of web based applications. Examples include Vertical Market Exchanges for credit accreditations and profile preferences; internet interface with wireless and cell phone devices; web site visitor cross-sell; and instant profiling of prospects over web or call center for tailored offerings to maximize loyalty.

 Back to top of page