System and method for fraud detection in a mobile device

Patent metadata

Field	Value
Publication number	US-10607226-B2
Application number	US-201514962365-A
Country	US
Kind code	B2
Filing date	Dec 8, 2015
Priority date	Apr 14, 2015
Publication date	Mar 31, 2020
Grant date	Mar 31, 2020

Abstract

A user device comprising: i) transmit path circuitry and receive path circuitry configured to communicate with a payment server; and ii) processing circuitry configured to control the transmit path circuitry and receive path circuitry. The processing circuitry is further configured to: a) receive a user input related to a payment process; b) calculate a risk score indicative of a likelihood of fraudulent activity associated with the payment process, wherein the risk score calculation is based on confidential information associated with the user that is stored on the user device; and c) transmit to the payment server a payment action and the risk score associated with the payment action without disclosing the confidential information. The confidential information comprises personally identifiable information and/or private information of the user. The processing circuitry calculates the risk score using a risk base model received from a model server.

First claim

What is claimed is:

1. A user device comprising:
transmit path circuitry and receive path circuitry configured to communicate with a payment server; and
processing circuitry configured to:
receive a user input related to a payment process,
calculate a risk score,
detect potentially fraudulent activity as a function of the risk score, wherein the risk score is indicative of a likelihood of fraudulent activity associated with the payment process, and wherein the risk score calculation is based on confidential information associated with a user that is stored on the user device and updatable parameters,
generate two sets of parameter update values for the risk score, wherein a first set of parameter update values relates to changes to make if fraudulent activity is not detected and a second set of parameter update values relates to changes to make if fraudulent activity is detected, and
transmit, to the payment server, a payment action and the risk score associated with the payment action without disclosing the confidential information.

2. The user device as set forth in claim 1 , wherein the confidential information comprises personally identifiable information of the user.

3. The user device as set forth in claim 1 , wherein the confidential information comprises private information of the user.

4. The user device as set forth in claim 1 , wherein the processing circuitry calculates the risk score using a risk base model received from a model server.

5. The user device as set forth in claim 4 , wherein the processing circuitry is further configured to transmit to the model server the two sets of parameter update values usable by the model server to improve accuracy of the risk base model.

6. The user device as set forth in claim 4 , wherein the risk base model is based on a neural network.

7. The user device as set forth in claim 4 , wherein the risk base model is based on a decision tree.

8. The user device as set forth in claim 1 , wherein the processing circuitry is further configured to transmit to the payment server a justification corresponding to the risk score.

9. For use in a user device, a method comprising:
communicating with a payment server;
receiving a user input related to a payment process;
calculating, by a mobile electronic device, a risk score;
detecting potentially fraudulent activity as a function of the risk score, wherein the risk score is indicative of a likelihood of fraudulent activity associated with the payment process at the mobile electronic device, and wherein the risk score calculation is based on confidential information associated with a user that is stored on the user device;
generating two set of parameter update values for the risk score, wherein a first set of parameter update values relates to changes to make if fraudulent activity is not detected and a second set of parameter update values relates to changes to make if fraudulent activity is detected; and
transmitting, to the payment server, a payment action and the risk score associated with the payment action without disclosing the confidential information.

10. The method as set forth in claim 9 , wherein the confidential information comprises personally identifiable information of the user.

11. The method as set forth in claim 9 , wherein the confidential information comprises private information of the user.

12. The method as set forth in claim 9 , wherein calculating the risk score comprises calculating the risk score using a risk base model received from a model server.

13. The method as set forth in claim 12 , further comprising:
transmitting to the model server the two sets of parameter update values usable by the model server to improve accuracy of the risk base model.

14. The method as set forth in claim 12 , wherein the risk base model is based on a neural network.

15. The method as set forth in claim 12 , wherein the risk base model is based on a decision tree.

16. The method as set forth in claim 9 , further comprising:
transmitting to the payment server a justification corresponding to the risk score.

17. A non-transitory computer readable medium including a plurality of instructions that, when executed by a processor, are configured to cause the processor to detect potentially fraudulent activity by a user of an electronic device including the processor by:
communicating with a payment server;
receiving a user input related to a payment process;
calculating a risk score;
detecting potentially fraudulent activity as a function of the risk score, wherein the risk score is indicative of a likelihood of fraudulent activity associated with the payment process, and wherein the risk score calculation is based on confidential information associated with the user that is stored on the user device and updatable parameters;
generating two sets of parameter update values for the risk score, wherein a first set of parameter update values relates to changes to make if fraudulent activity is not detected and a second set of parameter update values relates to changes to make if fraudulent activity is detected; and
transmitting to the payment server a payment action and the risk score associated with the payment action without disclosing the confidential information.

18. The non-transitory computer readable medium as set forth in claim 17 , wherein the confidential information comprises personally identifiable information of the user.

19. The non-transitory computer readable medium as set forth in claim 17 , wherein the confidential information comprises private information of the user.

20. The non-transitory computer readable medium as set forth in claim 17 , wherein the plurality of instructions is configured to cause the processor to:
calculate the risk score using a risk base model received from a model server; and
transmit to the model server the two sets of parameter update values usable by the model server to improve accuracy of the risk base model.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Patel Kunal M
• Kang Abraham J
• Kasman Bulent
• Ning Peng
• Grace Michael C

Classifications

• G06Q20/4016 Primary

  involving fraud or risk level assessment in transaction processing · CPC title

• G06Q20/322

  Aspects of commerce using mobile devices [M-devices] · CPC title

• G06F21/6245 Primary

  Protecting personal data, e.g. for financial or medical purposes · CPC title

• G06Q40/00

  Finance; Insurance; Tax strategies; Processing of corporate or income taxes · CPC title

• G06N3/02

  Neural networks · CPC title