Detecting fraudulent mobile payments
A method for processing an attempted payment made using a mobile device includes receiving information about the attempted payment, receiving data indicative of a behavior of a user of the mobile device at the time of the attempted payment, computing a likelihood that the attempted payment is fraudulent, based on a comparison of the behavior of the user to an historical behavior pattern of the user, and sending an instruction indicating how to proceed with the attempted payment, based on the likelihood.
1. A method, comprising:
receiving, by a server, information from a mobile payment application executing on a mobile device, wherein the information relates to a payment that the mobile payment application is attempting to make to a third party;
generating, by the server and using at least a first portion of the information, a first score indicative of how closely the payment matches an observed transaction pattern associated with a user of the mobile device;
generating, by the server and using at least a second portion of the information, a second score indicative of how closely a behavior of the user during the payment matches a first model, wherein the first model is synchronized, prior to the receiving, with a second model that is generated by the mobile device based on an observed behavioral pattern of the user;
generating, by the server and using at least a third portion of the information, a third score indicative of how risky the payment is;
aggregating, by the server, the first score, the second score, and the third score in order to generate a final score indicating a likelihood that the payment is fraudulent, wherein the aggregating comprises assigning different weights to each score of the first score, the second score, and the third score, and wherein the different weights change over time based on user feedback;
determining, by the server, whether the payment should proceed, based on the likelihood; and
sending, by the server, an instruction to the mobile payment application based on the determining, wherein the instruction instructs the mobile payment application to take an action with respect to the payment.
2. The method of claim 1 , at least some of the information is obtained using a sensor integrated in the mobile device.
3. The method of claim 1 , wherein the observed behavioral pattern of the user describes a pattern of interaction between the user and the mobile device while the user uses an application executing on the mobile device.
4. The method of claim 3 , wherein the pattern of interaction includes physical contact between the user and a touch screen of the mobile device.
5. The method of claim 4 , wherein the physical contact comprises at least one of: a swipe, a drag, a rotation, a flick, a pinch, a spread, or a tap.
6. The method of claim 1 , wherein the first model describes a mobility pattern of the user.
7. The method of claim 1 , wherein the first model describes a communication pattern of the user.
8. The method of claim 1 , wherein the first model describes a physiological pattern of the user.
9. The method of claim 1 , wherein the first model describes an environment in which the user tends to use the mobile device.
10. The method of claim 1 , further comprising:
periodically refining, by the server, the observed transaction pattern based on additional information provided by the mobile payment application, wherein the additional information related to additional payments that the mobile payment application attempts to make.
11. The method of claim 1 , wherein the third portion of the information comprises a usage by the user of a security mechanism during the payment.
12. The method of claim 1 , wherein the third portion of the information comprises a location from which the payment was initiated.
13. The method of claim 1 , further comprising:
receiving, by the server, user feedback from the mobile payment application, wherein the user feedback confirms or rejects at least some of the information; and
refining, by the server, the observed transaction pattern, wherein the refining is based on the user feedback.
14. The method of claim 13 , wherein the different weights change further over time based on the user feedback.
15. The method of claim 1 , further comprising:
prior to sending the instruction to the mobile payment application, sending, by the server, a request for authentication information to the mobile payment application.
16. The method of claim 1 , wherein the server and the mobile payment application communicate over an Internet Protocol network.
17. The method of claim 1 , where the mobile payment application communicates with a server to complete transactions for goods and services.
18. The method of claim 1 , wherein the first model is stored in a database that is accessible by the server.
19. The method of claim 18 , wherein the second model is stored on the mobile device.
20. A non-transitory computer readable storage device containing an executable program, where execution of the program causes a processor of a server in a communications network to perform steps comprising:
receiving, by a server, information from a mobile payment application executing on a mobile device, wherein the information relates to a payment that the mobile payment application is attempting to make to a third party;
generating, by the server and using at least a first portion of the information, a first score indicative of how closely the payment matches an observed transaction pattern associated with a user of the mobile device;
generating, by the server and using at least a second portion of the information, a second score indicative of how closely a behavior of the user during the payment matches a first model, wherein the first model is synchronized, prior to the receiving, with a second model that is generated by the mobile device based on an observed behavioral pattern of the user;
generating, by the server and using at least a third portion of the information, a third score indicative of how risky the payment is;
aggregating, by the server, the first score, the second score, and the third score in order to generate a final score indicating a likelihood that the payment is fraudulent, wherein the aggregating comprises assigning different weights to each score of the first score, the second score, and the third score, and wherein the different weights change over time based on user feedback;
determining, by the server, whether the payment should proceed, based on the likelihood; and
sending, by the server, an instruction to the mobile payment application based on the determining, wherein the instruction instructs the mobile payment application to take an action with respect to the payment.