US 12,406,260 B2
Fraud detection system, fraud detection device, fraud detection method, and program
Kyosuke Tomoda, Tokyo (JP); and Shuhei Ito, Tokyo (JP)
Assigned to RAKUTEN GROUP, INC., Tokyo (JP)
Appl. No. 17/596,419
Filed by RAKUTEN GROUP, INC., Tokyo (JP)
PCT Filed Dec. 11, 2020, PCT No. PCT/JP2020/046320
§ 371(c)(1), (2) Date Dec. 9, 2021,
PCT Pub. No. WO2022/123779, PCT Pub. Date Jun. 16, 2022.
Prior Publication US 2022/0351211 A1, Nov. 3, 2022
Int. Cl. G06Q 20/40 (2012.01); G06Q 20/32 (2012.01)
CPC G06Q 20/4016 (2013.01) [G06Q 20/3226 (2013.01); G06Q 20/3274 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A fraud detection system using a machine learning model for executing predetermined processing when a detection target is detected by using a detection device, the fraud detection system comprising at least one processor configured to:
determine, before the detection target is detected, whether a predetermined action has been performed by a user having a user terminal;
execute, when it is determined that the predetermined action has been performed, fraud detection on the user based on identification information stored in the user terminal, said fraud detection being executed by the machine learning model;
execute, when the detection target is detected, the predetermined processing based on an execution result of the fraud detection;
wherein the detection target is a code displayed on the user terminal,
wherein the detection device is used to detect the code, and is external to the user terminal,
wherein the predetermined action is a display action for displaying the code on the user terminal,
wherein the at least one processor is configured to determine, before the code is detected, whether the display action has been performed,
wherein, when it is determined that the display action has been performed, the at least one processor is configured to execute the fraud detection based on the identification information,
wherein, when the code is detected, the at least one processor is configured to execute the predetermined processing based on the execution result of the fraud detection;
wherein the machine learning model outputs a probability of the code being used by the user;
wherein the machine learning model outputs a fraud detection result, in real time, using the probability as an input;
wherein when fraud is detected, the user is not permitted to enter a facility; and
wherein when fraud is not detected, the user is permitted to enter the facility by opening an entrance gate to the facility.