US 11,775,947 B2
Method and system to predict ATM locations for users
Abdelkader M'Hamed Benkreira, Brooklyn, NY (US); Michael Mossoba, Great Falls, VA (US); and Joshua Edwards, Philadelphia, PA (US)
Assigned to Capital One Services, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Jun. 14, 2021, as Appl. No. 17/304,065.
Application 17/304,065 is a continuation of application No. 16/568,848, filed on Sep. 12, 2019, granted, now 11,037,115.
Prior Publication US 2021/0304166 A1, Sep. 30, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 40/00 (2023.01); G06Q 20/10 (2012.01); G06N 5/02 (2023.01)
CPC G06Q 20/1085 (2013.01) [G06N 5/02 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A method of routing customers to an automated teller machine (ATM), comprising:
retrieving, by a computing system associated with at least one financial institution having an ATM network, historical ATM transaction data for a user of the ATM network, wherein the historical ATM transaction data comprises a user identifier and attributes of a respective ATM associated with each historical ATM transaction executed by that user;
training, by the computing system using a machine learning module of the computing system, an individualized prediction model to learn and predict individualized ATM habits of the user based on training sets comprising the historical ATM transaction data of the user;
receiving, by the computing system from a client device associated with the user, a request to locate a target ATM, wherein the request comprises a constraint of a desired ATM, wherein the computing system is in electronic communication with an application installed on the client device for providing location services and real-time electronic dialog messaging services to the computing system, wherein the request is generated by the user using an interface generated on the client device by the application;
determining, by the computing system and based on the request, the user identifier associated with the user and a location of the client device;
parsing, by the computing system, the request using a natural language processor (NLP) to output a meaning of the request to the computing system in a format that the computing system can process using the individualized prediction model of the user;
identifying, by the computing system, a plurality of ATMs proximate the location of the client device based on the constraint, wherein the plurality of identified ATMs are selected from a group comprising financial institution ATMs and third party ATMs;
generating, by the computing system via the individualized prediction model of the user, a personalized ATM recommendation for the user by:
interfacing, by the computing system, with each respective ATM of the plurality of ATMs proximate to the location of the client device to identify attributes of each respective ATM; and
comparing, by the computing system, the attributes from each respective ATM of the plurality of ATMs to historical ATM usage statistics of the user; and
interfacing, with the user, via a dialog message session generated on the interface of the client device, to route the user of the client device to the target ATM from the plurality of ATMs based at least partially on the historical ATM usage statistics, wherein the dialog message session comprises a persistent session in which the computing system is configured to respond to questions from the user.