Network Risk: Reviewing the Feature in Transaction Risk API - Ekata
Unlock savings potential now with the ROI Fraud Calculator from Mastercard Identity! Calculate your savings instantly!

Unlock savings potential now with the ROI Fraud Calculator from Mastercard Identity! Calculate your savings instantly!

New Feature in Pro Insight: Network Risk

We researched and assessed the needs of our users to ensure this release would provide absolute value and are extremely excited that with the Network Risk panel, Pro Insight is now the only manual review solution that enables agents to investigate both the validity of identity elements as well as activity and usage patterns of those elements.

Ekata Network Risk Assessment

Understanding how identity elements are transacting online can be a crucial advantage for manual review teams in determining the riskiness of an identity or transaction. With fraud patterns evolving quickly during the global pandemic, knowing what your customers are doing or how their information is being used online is equally as important as knowing who they say they are. And that’s exactly what Network Risk provides— insight into the customer’s identity elements and how they are being used online. Fraud analysts can now make more accurate and confident decisions with Network Risk’s transaction-level intelligence and real-time answers.

How to Use Network Risk

Derived from the Ekata Identity Network, the Network Risk panel can be a useful secondary check when reviewing an identity. If Pro Insight’s identity data columns do not provide enough clarity for confident decision making, the panel can provide more detail on the consumer’s behavior, in addition to the validity of data points. Together, the two features validate identity elements as well as how they are used online.

Ekata Network Risk Assessment

Network Risk Features

transaction risk api Ekata

A. Risk Indicator at the top of the panel– green/low risk, red/high risk, and orange/uncertain.

  • This indicator is a machine learning prediction, powered by our Ekata Identity Network, that provides insight into how risky a digital interaction or transaction is based on activity patterns of the identity elements that are being used.

B. Strength Bars 

  • The number of strength bars indicates how much influence that signal had on the model.
  • The red/green colors represent the direction of that influence.
  • Red bars indicate a riskier leaning. Green bars represent a positive/less risky leaning.

C. Signals explain the indicator. (i.e. phone seen in 5 transactions in the last 30 days)

  • Our product and engineering teams tested hundreds of attributes that are within our Identity Network. After this testing, we chose a combination of 147 features (signals) based on their proven predictive value and trained our random forest machine learning model to look across these 147 signals, giving equal weight to each input type. Then, our model surfaces 1) the top signals based on their predictive power and 2) a Risk Indicator that “summarizes” the riskiness level of these features. In other words, the risk signals (i.e. email seen in 10 transactions) explain why the Risk Indicator is low, high, or uncertain.

D. Activity Patterns: 

  • Transaction velocity – the number of transactions an identity element has been used within a given time frame (ex. An email address used in 20 transactions in the last 60 days).
  • Merchant popularity – the number of merchants an identity element is seen with (ex. An email address used across 50 merchants in the last 90 days).
  • Attribute volatility – the number of times an identity element and transaction identity attributes change in transactions (ex. A phone number used with 100 shipping addresses in the last 30 days).
  • Attribute age – measures the reliability of an identity element based on its history in the Ekata Identity Network (ex. An email and IP first seen together 1 month ago).

 

The Ekata Identity Network

The Identity Network helps businesses identify good and bad customers during transactions by analyzing patterns of how their information is being used in digital interactions using transaction-level intelligence from more than 200M monthly queries.

Instead of using static data points based on historical transactions, the Identity Network provides insights based on predictive patterns. In addition, the Identity Network offers businesses dynamic decision-making, learning with each new transaction to provide answers in real time for determining fraud potential (not just static information used to identify credit risk, such as a credit score, social security number or date of birth).

Interested in how the Network Risk panel can help your business run more efficiently? Contact an Ekata fraud expert to schedule a Pro Insight demo today.

Start a Free Trial

See how Ekata can reduce fraud risk for your business, contact us for a Demo.