US 12,236,430 B2
Biometrics-infused dynamic knowledge-based authentication tool
Samuel Rapowitz, Roswell, GA (US); Viraj Chaudhary, Katy, TX (US); and Zviad Aznaurashvili, Reston, VA (US)
Assigned to Capital One Services, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Apr. 16, 2021, as Appl. No. 17/232,342.
Prior Publication US 2022/0335433 A1, Oct. 20, 2022
Int. Cl. G06Q 20/40 (2012.01); G06N 20/00 (2019.01); G10L 17/02 (2013.01); G10L 17/04 (2013.01)
CPC G06Q 20/40145 (2013.01) [G06N 20/00 (2019.01); G10L 17/02 (2013.01); G10L 17/04 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
training, by a computing device, a first machine learning model to identify one or more memorable transactions based on features of the one or more memorable transactions;
receiving, by the computing device voice data, corresponding to a user, that comprises a recorded voice sample of the user speaking;
defining, by the computing device, a plurality of characteristics of the user by processing the voice data to identify one or more speech patterns of the user;
training, based on the processed voice data, a second machine learning model to recognize the one or more speech patterns of the user;
receiving, by the computing device and from a second computing device, a request for authorization to perform a current financial transaction;
receiving, by the computing device, data associated with a plurality of prior financial transactions of the user;
identifying, by the computing device, one or more prior financial transactions of the plurality of prior financial transactions by:
providing, as input to the first machine learning model, the plurality of prior financial transactions; and
receiving, as output from the first machine learning model, an indication of the one or more prior financial transactions;
generating, by the computing device, one or more challenge questions based on the one or more prior financial transactions of the user;
providing the one or more challenge questions in response to the request for authorization to perform the current financial transaction;
receiving, by the computing device, one or more audible responses to each of the one or more challenge questions;
determining, by the computing device, using the second machine learning model, and based on second speech patterns of the one or more audible responses, a likelihood that the user provided the one or more audible responses;
comparing the determined likelihood to a predetermined threshold; and
determining whether to grant the request for authorization to perform the current financial transaction based on the determined likelihood exceeding the predetermined threshold or to deny the request for authorization to perform the current financial transaction based on the determined likelihood not exceeding the predetermined threshold.