minFraud Release Notes - 2016 Archive

minFraud's is_postal_in_city available for non-US locations

minFraud's is_postal_in_city, cityPostalMatch, and shipCityPostalMatch outputs are available for non-US locations.

Disposition Outputs for minFraud Score, Insights, and Factors

The responses for minFraud Score, Insights, and Factors may now include a disposition object. This object contains information about custom rules that were applied to the request.

New /credit_card/token Input

We have added a new /credit_card/token input to the minFraud Score, minFraud Insights, and minFraud Factors request bodies. The token must uniquely identify the credit card and must be different from the account number.

New event types for minFraud requests

We have added two new event types to the /event/type field for all minFraud requests, email_change and password_reset.

More proxy scores

We will be returning more proxyScores between 0 and 1.8 and 1.8 and 3. Such scores typically represent “medium” risk transactions which have some risk of fraud but may be legitimate. The higher the proxyScore, the riskier the transaction.

The proxyScore is a score from 0.00-4.00 indicating the likelihood that the user's IP address is high risk.

This change has also been implemented for the IP Risk Score provided with minFraud Score, minFraud Insights, and minFraud Factors. The IP Risk Score is a score from 0.01 to 99 indicating the likelihood that the user's IP address is high risk. We will return more positive IP Risk Scores between 0 and 45 and 45 and 75. Such scores typically represent “medium” risk transactions which have some risk of fraud but may be legitimate. The higher the IP Risk Score, the riskier the transaction.

New Device Field

/device/confidence and /device/last_seen have been added to the minFraud Insights and minFraud Factors response. These fields are only available for customers using our Device Tracking Add-on. Please see our API documentation for more information.

Deprecated Fields and New Fields

credits_remaining is deprecated in the minFraud Score, minFraud Insights, and minFraud Factors response. It will be removed before the final release. It has been replaced with queries_remaining, which also returns the number of queries available. An additional field, funds_remaining, containing the US dollar value of the funds remaining will be added before the final release.

Order Amount Now Used in Risk Score

All minFraud services now include an updated machine learning algorithm that takes order amount into consideration when calculating the risk score.

Merchants may want to update their rules if the rule is a combination of risk score and order amount. For example, if a custom rule rejects transactions with a high order amount and a risk score above 10, the merchant may want to adjust the risk score to a higher number, as the high order amount itself may cause the risk score to increase.

In general, orders $500 or above may have a higher risk score than before.

New minFraud Insights Outputs

The following outputs have been added to minFraud Insights:

• /credit_card/brand – The card brand, such as “Visa”, “Discover”, “American Express”, etc.
• /credit_card/type – A string representing the card's type. The possible values are charge, credit, or debit.
• /device/id – A UUID that MaxMind uses for the device associated with this IP address. This is only available if you are using the Device Tracking Add-on.
• /email/is_free – This field is true if MaxMind believes that this email is hosted by a free email provider such as Gmail or Yahoo! Mail.
• /email/is_high_risk – This field is true if MaxMind believes that this email is likely to be used for fraud.

Also, for both minFraud Score and minFraud Insights, the input property in the warning objects has been removed and replaced with an input_pointer property. This new property is a JSON pointer to the value in the input object causing the warning.

All of our client APIs have been updated to reflect these changes.

For more information on these outputs, please see our minFraud Score and minFraud Insights API documentation.