IP Library › Granted Patent US 11,587,078
Granted Patent B2
US 11,587,078 · App. 17/017,860 · Granted Feb 21, 2023

System, method, and computer program product for predicting payment transactions using a machine learning technique based on merchant categories and transaction time data

Inventors: Amitava Dutta (Singapore, SG); April Pabale Vergara (Singapore, SG); Suresh Krishna Vaidyanathan (Dublin, CA)
Assignee: Visa International Service Association
G06Q20/40G06N5/048G06N7/005G06N20/00G06Q20/20G06Q20/34G06Q20/405G06Q20/4015G06Q20/4016G06Q30/0254G06Q30/0261G07F7/0609G06Q10/067G06Q30/0215G06Q30/0238
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,587,078
App. No.
17/017,860
Granted
Feb 21, 2023
Kind
B2
Abstract

Provided is a computer-implemented method for predicting payment transactions using a machine learning technique that includes receiving transaction data, generating a categorical transaction model based on the transaction data, determining a plurality of prediction scores including determining, for one or more users, a prediction score in each merchant category of a plurality of merchant categories for each predetermined time segment of a plurality of predetermined time segments, where a respective prediction score includes a prediction of whether a user will conduct a payment transaction in a merchant category at a time associated with a predetermined time segment associated with the respective prediction score, determining a recommended merchant category and a recommended predetermined time segment of at least one offer, generating the at least one offer, and communicating the at least one offer to the one or more users. A system and computer program product are also disclosed.

Claims (63)

1. A computer-implemented method for predicting payment transactions using a machine learning technique, the method comprising:

receiving, with at least one processor, historical transaction data, wherein the historical transaction data is associated with a plurality of historical payment transactions in a plurality of merchant categories, wherein the plurality of historical payment transactions involve a plurality of users;

generating, with at least one processor, a categorical transaction model based on the transaction data, wherein the categorical transaction model comprises a model designed to receive, as an input, transaction data associated with a plurality of payment transactions, and provide, as an output, a prediction as to whether a user will conduct a transaction in a merchant category of a plurality of merchant categories and in a predetermined future time segment of a plurality of predetermined future time segments, wherein generating the categorical transaction model comprises:

processing the historical transaction data to obtain training data for the dominant account profile classification model, wherein processing the historical transaction data comprises:

determining a set of transaction variables based on the historical transaction data;

changing the historical transaction data into a format to be analyzed to generate the categorical transaction model; and

generating the categorical transaction model based on training the categorical transaction model using the set of transaction variables; and

determining, with at least one processor, a plurality of prediction scores for one or more users based on the categorical transaction model and additional transaction data, wherein determining the plurality of prediction scores comprises:

determining, for the one or more users of the plurality of users, a prediction score in each merchant category of the plurality of merchant categories for each predetermined future time segment of the plurality of predetermined future time segments, wherein a respective prediction score comprises a prediction of whether the one or more users will conduct a payment transaction in a merchant category of the plurality of merchant categories associated with the respective prediction score, at a future time associated with a predetermined future time segment of the plurality of predetermined future time segments associated with the respective prediction score, wherein the predetermined future time segment comprises a predetermined future time of day segment and a predetermined future day of a week segment, and wherein the predetermined future time of day segment is a predetermined future time of day segment of a plurality of predetermined future time of day segments, wherein the predetermined future day of a week segment is a predetermined future day of a week segment of a plurality of predetermined future day of a week segments, wherein the plurality of predetermined future time of day segments comprises at least four future time of day segments, and wherein the plurality of predetermined future time of day segments comprises a future time of day segment associated with night, a future time of day segment associated with evening, and a future time of day segment associated with morning; and

determining a recommended merchant category for each predetermined future time of day segment of the plurality of predetermined future time of day segments and one or more recommended predetermined future time segments for each offer of a plurality of offers based on the prediction scores of the one or more users;

generating the plurality of offers based on the recommended merchant categories and the one or more recommended predetermined future time of day segments of the plurality of offers; and

communicating one or more offers of the plurality of offers to the one or more users at a time of day corresponding to the predetermined future time of day segment of each offer based on generating the plurality of offers.

2. The computer-implemented method of claim 1 , wherein the plurality of predetermined future day of the week segments comprises at least two predetermined future day of the week segments.

3. The computer-implemented method of claim 1 , wherein the recommended merchant categories and the one or more recommended predetermined future time of day segments of each offer correspond to a merchant category and a predetermined future time of day segment, respectively, that are determined to be associated with a prediction score for the one or more users that satisfies a threshold prediction score, and wherein the threshold prediction score comprises a highest prediction score for the one or more users in each merchant category of the plurality of merchant categories and the one or more predetermined future time of day segments of the plurality of predetermined future time of day segments.

4. The computer-implemented method of claim 3 , wherein the threshold prediction score comprises 10% percent of a plurality of highest prediction scores for the plurality of users in the one or more merchant categories of the plurality of merchant categories and the one or more predetermined future time segments of the plurality of predetermined time segments.

5. The computer-implemented method of claim 1 , further comprising:

determining location data associated with a location of the one or more users; and

determining merchant identity data based on the location data associated with the location of the one or more users; and

wherein communicating the one or more offers of the plurality of offers to the one or more users at the time of day corresponding to the predetermined future time of day segment of each offer comprises:

communicating an offer to the one or more users based on the merchant identity data.

6. A system for predicting payment transactions using a machine learning technique, the system comprising:

at least one processor programmed or configured to:

receive historical transaction data, wherein the historical transaction data is associated with a plurality of historical payment transactions in a plurality of merchant categories, wherein the plurality of historical payment transactions involve a plurality of users;

generate a categorical transaction model based on the transaction data, wherein the categorical transaction model comprises a model designed to receive, as an input, transaction data associated with a plurality of payment transactions, and provide, as an output, a prediction as to whether a user will conduct a transaction in a merchant category of a plurality of merchant categories and in a predetermined future time segment of a plurality of predetermined future time segments, wherein, when generating the categorical transaction model, the at least one processor is programmed or configured to:

process the historical transaction data to obtain training data for the dominant account profile classification model, wherein, when processing the transaction data, the at least one processor is programmed or configured to:

determine a set of transaction variables based on the historical transaction data;

change the historical transaction data into a format to be analyzed to generate the categorical transaction model; and

generate the categorical transaction model based on training the categorical transaction model using the set of transaction variables and the historical transaction data; and

determine a plurality of prediction scores for one or more users based on the categorical transaction model and additional transaction data, wherein when determining the plurality of prediction scores, the at least one processor is programmed or configured to:

determine, for the one or more users of the plurality of users, a prediction score in each merchant category of the plurality of merchant categories for each predetermined future time segment of the plurality of predetermined future time segments, wherein a respective prediction score comprises a prediction of whether the one or more users will conduct a payment transaction in a merchant category of the plurality of merchant categories associated with the respective prediction score, at a time associated with a predetermined future time segment of the plurality of predetermined future time segments associated with the respective prediction score, wherein the predetermined future time segment comprises a predetermined future time of day segment and a predetermined future day of a week segment, and wherein the predetermined future time of day segment is a predetermined future time of day segment of a plurality of predetermined future time of day segments, wherein the predetermined future day of a week segment is a predetermined future day of a week segment of a plurality of predetermined future day of a week segments, wherein the plurality of predetermined future time of day segments comprises at least four future time of day segments, and wherein the plurality of predetermined future day of the week segments comprises at least two predetermined future day of the week segments, wherein the plurality of predetermined future time of day segments comprises at least four future time of day segments, and wherein the plurality of predetermined future time of day segments comprises a future time of day segment associated with night, a future time of day segment associated with evening, and a future time of day segment associated with morning; and

determine a recommended merchant category for each predetermined future time of day segment of the plurality of predetermined future time of day segments and one or more recommended predetermined future time segments for each offer of a plurality of offers based on the prediction scores of the one or more users;

generate the plurality of offers based on the recommended merchant categories and the one or more recommended predetermined future time of day segments of the plurality of offers; and

communicate one or more offers of the plurality of offers to the one or more users at a time of day corresponding to the predetermined future time of day segment of each offer based on generating the plurality of offers.

7. The system of claim 6 , wherein the at least one processor, when generating the categorical transaction model, is programmed or configured to:

generate the categorical transaction model based on at least one machine learning technique.

8. The system of claim 6 , wherein the at least one processor is further programmed or configured to:

determine location data associated with a location of the one or more users; and

determine merchant identity data based on the location data associated with the location of the one or more users; and

wherein, when communicating the one or more offers of the plurality of offers to the one or more users at the time of day corresponding to the predetermined future time of day segment of each offer, the at least one processor is programmed or configured to:

communicate an offer to the one or more users based on the merchant identity data.

9. The system of claim 6 ,

wherein the plurality of predetermined future day of the week segments comprises at least two predetermined future day of the week segments.

10. A computer program product for predicting payment transactions using a machine learning technique, the computer program product comprising at least one non-transitory computer-readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to:

receive historical transaction data, wherein the historical transaction data is associated with a plurality of historical payment transactions in a plurality of merchant categories, wherein the plurality of historical payment transactions involve a plurality of users;

generate a categorical transaction model based on the historical transaction data, wherein the categorical transaction model comprises a model designed to receive, as an input, transaction data associated with a plurality of payment transactions, and provide, as an output, a prediction as to whether a user will conduct a transaction in a merchant category of a plurality of merchant categories and in a predetermined future time segment of a plurality of predetermined future time segments, wherein the one or more instructions that cause the at least one processor to generate the categorical transaction model, cause the at least one processor to:

process the historical transaction data to obtain training data for the dominant account profile classification model, wherein the one or more instructions that cause the at least one processor to process the transaction data, cause the at least one processor to:

determine a set of transaction variables based on the historical transaction data;

change the historical transaction data into a format to be analyzed to generate the categorical transaction model; and

generate the categorical transaction model based on training the categorical transaction model using the set of transaction variables and the historical transaction data; and

determine a plurality of prediction scores for one or more users based on the categorical transaction model and additional transaction data, wherein the one or more instructions that cause the at least one processor to determine a plurality of prediction scores, cause the at least one processor to:

determine, for the one or more users of the plurality of users, a prediction score in each merchant category of the plurality of merchant categories for each predetermined future time segment of the plurality of predetermined future time segments, wherein a respective prediction score comprises a prediction of whether the one or more users will conduct a payment transaction in a merchant category of the plurality of merchant categories associated with the respective prediction score, at a time associated with a predetermined future time segment of the plurality of predetermined future time segments associated with the respective prediction score, wherein the predetermined future time segment comprises a predetermined future time of day segment and a predetermined future day of a week segment, and wherein the predetermined future time of day segment is a predetermined future time of day segment of a plurality of predetermined future time of day segments, wherein the predetermined future day of a week segment is a predetermined future day of a week segment of a plurality of predetermined future day of a week segments, wherein the plurality of predetermined future time of day segments comprises at least four future time of day segments, and wherein the plurality of predetermined future day of the week segments comprises at least two predetermined future day of the week segments, wherein the plurality of predetermined future time of day segments comprises at least four future time of day segments, and wherein the plurality of predetermined future time of day segments comprises a future time of day segment associated with night, a future time of day segment associated with evening, and a future time of day segment associated with morning; and

determine a recommended merchant category for each predetermined future time of day segment of the plurality of predetermined future time of day segments and one or more recommended predetermined future time segments for each offer of a plurality of offers based on the prediction scores of the one or more users;

generate the plurality of offers based on the one or more recommended merchant categories and the one or more recommended predetermined future time of day segments of the at least one offer; and

communicate one or more offers of the plurality of offers to the one or more users at a time of day corresponding to the predetermined future time of day segment of each offer based on generating the plurality of offers.

11. The computer program product of claim 10 , wherein the recommended merchant categories and the one or more recommended predetermined future time of day segments of each offer correspond to a merchant category and a predetermined future time of day segment, respectively, that are determined to be associated with a prediction score for the one or more users that satisfies a threshold prediction score, and

wherein the threshold prediction score comprises a highest prediction score for the one or more users in each merchant category of the plurality of merchant categories and the one or more predetermined future time of day segments of the plurality of predetermined future time of day segments.

12. The computer program product of claim 10 , wherein the one or more instructions, when executed by the at least one processor, further cause the at least one processor to:

determine location data associated with a location of the one or more users; and

determine merchant identity data based on the location data associated with the location of the one or more users; and

wherein, the one or more instructions that cause the at least one processor to communicate the one or more offers of the plurality of offers to the one or more users at the time of day corresponding to the predetermined future time of day segment of each offer, cause the at least one processor to:

communicate an offer to the one or more users based on the merchant identity data.

13. The computer program product of claim 10 ,

wherein the plurality of predetermined future day of the week segments comprises at least two predetermined future day of the week segments.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 11, 2020
From: DUTTA, AMITAVA; VERGARA, APRIL PABALE; VAIDYANATHAN, SURESH KRISHNA
To: VISA INTERNATIONAL SERVICE ASSOCIATION
Reel/Frame 053757/0761 →
Continuity (2)
Continuation 15696447 · Sep 6, 2017
Related Publication 20200410490A1 · Dec 31, 2020