US 11,748,727 B2
Systems and methods for a user interface for making recommendations
David Gabriele, New York, NY (US); George Bergeron, Falls Church, VA (US); Adam Vukich, Alexandria, VA (US); Vu Nguyen, Pittsburg, CA (US); Andrea Montealegre, Arlington, VA (US); Ljubica Chatman, New York, NY (US); Mykhaylo Bulgakov, Arlington, VA (US); Joshua Edwards, Philadelphia, PA (US); and Abdelkader Benkreira, Washington, DC (US)
Assigned to CAPITAL ONE SERVICES, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Jun. 17, 2020, as Appl. No. 16/904,115.
Prior Publication US 2021/0398100 A1, Dec. 23, 2021
Int. Cl. G06Q 20/20 (2012.01); G06Q 20/38 (2012.01); G06Q 20/22 (2012.01); G06N 20/00 (2019.01)
CPC G06Q 20/204 (2013.01) [G06N 20/00 (2019.01); G06Q 20/227 (2013.01); G06Q 20/387 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system, comprising:
an application programming interface for making recommendations to a user at a point of sale to display a recommended payment method from a set of payment methods associated with the user, wherein the application programming interface stores information about the user and one or more predetermined recommended payment methods associated with one or more web sites in a cache;
a database that stores past transactions for each payment method in the set of payment methods and a set of generic transactions; and
a server that is configured to:
receive a payment request from the point of sale;
determine that the payment request attempts to use one of the payment methods associated with the user by comparing payment information included in the payment request with payment information associated with each of the respective payment methods;
identify the payment request payment method based on comparison of payment information included in the payment request with payment information associated with each of the respective payment methods;
train a predictive model using the information stored in the cache;
filter one or more past transactions in the database for each payment method based on one or more expired time periods of potential benefits to the user to yield one or more filtered past transactions and one or more unwanted results, wherein:
the one or more filtered past transactions correspond to one or more past transactions selected by the filter, and
the one or more unwanted results correspond to one or more past transactions removed by the filter;
generate payment analytics by using the predictive model to analyze the one or more filtered past transactions in the database for each payment method;
determine potential benefits available to the user;
determine the recommended payment method by using the payment analytics and maximizing the potential benefits available to the user;
process the payment request using the payment request payment method if the recommended payment method is the same as the payment request payment method; and
transmit to the point of sale an indication to be displayed to the user by the point of sale that the recommended payment method is different than the payment request payment method if the recommended payment method is the different than the payment request payment method;
receive, from the point of sale, a user choice of a payment method from the set of payment methods;
retrieve, from the database, one or more generic transactions from the set of generic transactions for each of the unwanted results; and
retrain the predictive model based on the one or more filtered past transactions, the one or more generic transactions, and the user choice.