Mitigation of fraudulent transactions conducted over a network
The subject matter disclosed herein provides methods for detecting and mitigating fraudulent transactions conducted over a network. This method can access one or more data values associated with an online transaction between a first party and a second party. The first party can initiate the online transaction. The data values can characterize an identity of the first party and the second party. A score representing a likelihood that the online transaction is fraudulent can be generated. The generation of the score can be based on one or more models for detecting fraud using the data values. The score can be compared to one or more threshold scores. The online transaction can be authorized or rejected based on the comparing. Related apparatus, systems, techniques, and articles are also described.
1. A non-transitory computer-readable media comprising computer-readable instructions stored thereon that when executed by one or more processors cause the one or more processors to:
receive one or more first data values associated with a buyer, one or more second data values associated with a recipient, and one or more third data values associated with payment information of the buyer, wherein the one or more first data values, the one or more second data values, and the one or more third data values are received in association with a request from the buyer during an online transaction to buy a good or service for the recipient;
determine a relationship between the buyer and the recipient based on the one or more first data values, the one or more first data values including one or more characteristics of a message associated with the online transaction;
compute an overall risk score based on at least one of the one or more first data values, the one or more second data values, or the one or more third data values and one or more fraud detection models, wherein the one or more fraud detection models comprise verifying the relationship between the buyer and the recipient by:
identifying a set of attributes of the buyer and the recipient based on the one or more first data values, the one or more second data values, and the one or more third data values;
searching the one or more separate sources to determine information about the buyer and the recipient;
comparing the set of attributes to the information about the buyer and the recipient to verify the set of attributes; and
assigning a score contribution of the fraud detection model to the overall risk score based on a number of the set of attributes that are verified based on the information determined from the one or more sources,
wherein the overall risk score represents a likelihood that the online transaction is fraudulent, and wherein the overall risk score is computed by applying a weight to each respective fraud detection model of the one or more fraud detection models, multiplying an output of the respective fraud detection model by the assigned weight to obtain a score contribution of the respective fraud detection model, and summing the score contributions of the one or more fraud detection models to compute the overall risk score;
compare the overall risk score with a threshold score; and
authorize or deny the online transaction based upon the comparison.
2. The non-transitory computer-readable media of claim 1 , wherein the one or more processors further execute computer-readable instructions to deny the online transaction based on the overall risk score being greater than the threshold score, and wherein the one or more processors further executes computer-readable instructions to deny the online transaction before the buyer completes the online transaction.
3. The non-transitory computer-readable media of claim 1 , wherein the one or more first data values comprise data associated with at least one of an identity of the buyer, an IP address of a computing device from which the online transaction is being conducted, a type of browser being used on the computing device for the online transaction, an operating system running on the computing device, or a geographic location of the computing device, wherein the one or more second data values comprise data associated with at least one of a name of the recipient, an e-mail or mailing address of the recipient, a phone number of the recipient, or a relationship between the buyer and the recipient, and wherein the one or more third data values comprise data associated with the identity of the buyer, the phone number of the buyer, the mailing address of the buyer, the e-mail address of the buyer, a billing address of the buyer, a payment card number, a card verification value, or a payment card expiration date.
4. The non-transitory computer-readable media of claim 1 , wherein the check determines whether any of the one or more first data values, the one or more second data values, and the one or more third data values were previously used to conduct fraudulent transactions, and wherein the one or more processors further execute computer-readable instructions to pass the check upon determining that none of the one or more first data values, the one or more second data values, and the one or more third data values match any value on a blacklist.
5. The non-transitory computer-readable media of claim 1 , wherein a fraud detection model of the one or more fraud detection models is based on an e-mail address of the buyer or the recipient, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on at least a subset of a legitimacy of the e-mail address, a domain name of the e-mail address, and a string of characters in the e-mail address.
6. The non-transitory computer-readable media of claim 1 , wherein a fraud detection model of the one or more fraud detection models is based on an IP address of a computing device from which the online transaction is conducted, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on a number of transactions from the IP address in a predetermined time period.
7. The non-transitory computer-readable media of claim 1 , wherein a fraud detection model of the one or more fraud detection models is based on a merchant associated with the online transaction, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on one or more statistics associated with the merchant or a line of business of the merchant.
8. The non-transitory computer-readable media of claim 1 , wherein the threshold score comprises a low value and a high value, and wherein the one or more processors further execute computer-readable instructions to:
deny the online transaction upon determining that the overall risk score is greater than the high value;
approve the online transaction upon determining that the overall risk score is lower than the low value; or
request a manual review of the online transaction upon determining that the overall risk score is greater than the low value and less than the high value before denying or approving the online transaction.
9. A system comprising:
a memory having computer-readable instructions stored thereon; and
one or more processors that execute the computer-readable instructions to:
receive one or more first data values associated with a buyer, one or more second data values associated with a recipient, and one or more third data values associated with payment information of the buyer, wherein the one or more first data values, the one or more second data values, and the one or more third data values are received in association with a request from the buyer during an online transaction to buy a good or service for the recipient;
determine a relationship between the buyer and the recipient based on the one or more first data values, the one or more first data values including one or more characteristics of a message associated with the online transaction;
compute an overall risk score based on at least one of the one or more first data values, the one or more second data values, or the one or more third data values and one or more fraud detection models, wherein the one or more fraud detection models comprise verifying the relationship between the buyer and the recipient by:
identifying a set of attributes of the buyer and the recipient based on the one or more first data values, the one or more second data values, and the one or more third data values;
searching the one or more separate sources to determine information about the buyer and the recipient;
comparing the set of attributes to the information about the buyer and the recipient to verify the set of attributes; and
assigning a score contribution of the fraud detection model to the overall risk score based on a number of the set of attributes that are verified based on the information determined from the one or more sources,
wherein the overall risk score represents a likelihood that the online transaction is fraudulent, and wherein the overall risk score is computed by applying a weight to each respective fraud detection model of the one or more fraud detection models, multiplying an output of the respective fraud detection model by the assigned weight to obtain a score contribution of the respective fraud detection model, and summing the score contributions of the one or more fraud detection models to compute the overall risk score;
compare the overall risk score with a threshold score; and
authorize or deny the online transaction based upon the comparison.
10. The system of claim 9 , wherein the one or more processors further execute computer-readable instructions to deny the online transaction based on the overall risk score being greater than the threshold score, and wherein the one or more processors further executes computer-readable instructions to deny the online transaction before the buyer completes the online transaction.
11. The system of claim 9 , wherein the one or more first data values comprise data associated with at least one of an identity of the buyer, an IP address of a computing device from which the online transaction is being conducted, a type of browser being used on the computing device for the online transaction, an operating system running on the computing device, or a geographic location of the computing device, wherein the one or more second data values comprise data associated with at least one of a name of the recipient, an e-mail or mailing address of the recipient, a phone number of the recipient, or a relationship between the buyer and the recipient, and wherein the one or more third data values comprise data associated with the identity of the buyer, the phone number of the buyer, the mailing address of the buyer, the e-mail address of the buyer, a billing address of the buyer, a payment card number, a card verification value, or a payment card expiration date.
12. The system of claim 9 , wherein the check determines whether any of the one or more first data values, the one or more second data values, and the one or more third data values were previously used to conduct fraudulent transactions, and wherein the one or more processors further execute computer-readable instructions to pass the check upon determining that none of the one or more first data values, the one or more second data values, and the one or more third data values match any value on a blacklist.
13. The system of claim 9 , wherein a fraud detection model of the one or more fraud detection models is based on an e-mail address of the buyer or the recipient, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on at least a subset of a legitimacy of the e-mail address, a domain name of the e-mail address, and a string of characters in the e-mail address.
14. The system of claim 9 , wherein a fraud detection model of the one or more fraud detection models is based on an IP address of a computing device from which the online transaction is conducted, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on a number of transactions from the IP address in a predetermined time period.
15. The system of claim 9 , wherein a fraud detection model of the one or more fraud detection models is based on a merchant associated with the online transaction, and wherein the one or more processors further execute computer-readable instructions to assign a score contribution of the fraud detection model to the overall risk score based on one or more statistics associated with the merchant or a line of business of the merchant.
16. The system of claim 9 , wherein the threshold score comprises a low value and a high value, and wherein the one or more processors further execute computer-readable instructions to:
deny the online transaction upon determining that the overall risk score is greater than the high value;
approve the online transaction upon determining that the overall risk score is lower than the low value; or
request a manual review of the online transaction upon determining that the overall risk score is greater than the low value and less than the high value before denying or approving the online transaction.