By: Kennis Wong
Data is the very core of fraud detection. We are constantly seeking new and mining existing data sources that give us more insights into consumers’ fraud and identity theft risk. Here is a way to categorize the various data sources.
Account level – When organizations detect fraud, naturally they leverage the data in-house. This type of data is usually from the individual account activities such as transactions, payments, locations or types of purchases, etc. For example, if there’s a purchase $5000 at a dry cleaner, the transaction itself is suspicious enough to raise a red flag.
Customer level – Most of the times we want to see a bigger picture than only at the account level. If the customer also has other accounts with the organization, we want to see the status of those accounts as well. It’s not only important from a fraud detection perspective, but it’s also important from a customer relationship management perspective.
Consumer level – As Experian Decision Analytics’ clients can attest, sometimes it’s not sufficient to look only at the data within an organization but also to look at all the financial relationships of the consumer. For example, in the situation of bust out fraud or first-party fraud, if you only look at the individual account, it wouldn’t be clear whether a consumer has truly committed the fraud. But when you look at the behavior of all the financial relationships, then the picture becomes clear.
Identity level – Fraud detection can go into the identity level. What I mean is that we can tie a consumer’s individual identity elements with those of other consumers to discover hidden inconsistencies and relationships. For example, we can observe the use of the same SSN across different applications and see if the phones or addresses are the same. In the account management environment, when detecting existing account fraud or account takeover, this level of linkage is very useful as more data becomes available after the account is open.
