At this week’s Fraud Force Summit, we proudly announced iovationScore, an adaptive score that uses sophisticated machine learning algorithms to predict the trustworthiness or riskiness of online transactions to distinguish great customers from fraudsters.
Machine learning, which is a form of artificial intelligence, has been a hot topic for several years now. From image recognition to speech recognition, and even for the search for extraterrestrial life, many disciplines make use of machine learning algorithms. Online fraud detection is no different, and can also benefit from machine learning technologies.
iovation has utilized machine learning algorithms in recent years for improving our device recognition technology through fuzzy pattern matching, as well as for determining the risk profiles of Internet-connected devices. However, iovationScore represents a major technological advancement in our machine learning capabilities. Years of research were invested in developing iovationScore, which can effectively predict the riskiness or trustworthiness of an online transaction.
Complex patterns from device, contextual, behavioral, and transactional attributes are analyzed to determine level of risk. iovationScore is “trained” from confirmed incidents of fraud reported by our subscriber network. More than 30M incidents of fraud have been reported by our global network of 3,500 fraud and security analysts.
From our perspective, the “wow factor” is that we can make these predictions even for transactions that involve devices and users that you’ve never seen before.
iovationScore provides our customers a long list of benefits. Here are a few of the bigger ones:
- First, it can be used to supplement our customers’ existing risk rules, which are usually based on locally detected patterns of fraud. iovationScore expands this and looks for global fraud patterns, even those that are too subtle or complex for a fraud analyst to manually spot. In addition, our customers can use iovationScore to prioritize the transactions that require manual review. Transactions with high scores (which represent trustworthy transactions) can be automatically approved, whereas transactions with low scores can be moved to the top of the review queue to be scrutinized more closely.
- Second, our customers can use iovationScore to better their customer experience and be more competitive by offering special promotions, incentives, and/or increased spending limits to customers with the highest scores. In effect, you can use iovationScore to incentivize new good customers to do business with you and existing good customers to stay with you, all while minimizing the risk of online fraud. High scores can also be used as a basis for accelerating orders through an order or payment approval system, which also improves the experience for your best customers.
Stay tuned for upcoming webinars that will provide more details about iovationScore. If you don’t want to wait for the webinar, simply contact your iovation client services manager to learn more.
This has been quite the week so far, with tons of interest from our client base about iovationScore. It’s safe to say that we’re already looking forward to the month ahead, all the way to next year’s Fraud Force Summit!