Scalable architecture for managing transactions

Patent metadata

Field	Value
Publication number	US-11250501-B2
Application number	US-201816107480-A
Country	US
Kind code	B2
Filing date	Aug 21, 2018
Priority date	Aug 21, 2018
Publication date	Feb 15, 2022
Grant date	Feb 15, 2022

Abstract

Provided, in an aspect, is a system for managing transactions from a financial services computer network. The system has a scalable architecture, and includes a database, one or more processors, a plurality of experience matchers, and a transaction hub. Each experience matcher is associated with a different experience set type, and can receive a transaction list and return an identified experience set. The transaction hub can communicate with multiple experience matchers, and can receive experience sets from the matchers after sending them a list of transactions.

First claim

What is claimed is:

1. A system for managing transactions from a financial services computer network, the system comprising:
a database;
one or more processors;
a plurality of experience matchers, each of the experience matchers being associated with a different experience set type, each of the experience matchers being configured for execution by the one or more processors to:
receive a list of transactions;
identify, by searching within the list of transactions, an experience set having an experience set type;
identify a set of the transactions as being encompassed by the experience set based on one or more of a location or time associated with the transaction; and
return the identified experience set; and

a transaction hub configured for execution by the one or more processors to:
receive a transaction stream from the financial services computer network;
send the list of transactions generated from the transaction stream to the plurality of experience matchers;
receive a first identified experience set from a first of the experience matchers and a second identified experience set from a second of the experience matchers based on the list of transactions;
in response to the first identified experience set and the second identified experience set sharing at least one of the transactions, analyze, by a machine learning model the shared transaction, the first identified experience set, the second identified experience set, and historical transactions and experience sets of the user, the machine learning model being trained to prioritize allocation of transactions to experience sets on a per user basis by learning a hierarchy of experience sets;
choose, by the machine learning model, one of the first identified experience set or the second identified experience set to be associated with the shared transaction;
assign the shared transaction to the chosen experience set; and
store the chosen experience set in the database,
wherein the one or more processors are configured to provide transaction details associated with the chosen experience set for display to a user in a graphical user interface;
wherein the one or more processors continuously update the machine learning model by retraining the machine learning model with the assignment of the shared transaction and the chosen experience set.

2. The system of claim 1 , wherein each of the experience matchers is configured for execution by the one or more processors to further determine, and return to the transaction hub, a match score for an identified experience set.

3. The system of claim 2 , wherein the transaction hub is configured for execution by the one or more processors to further prioritize allocation of a transaction into an identified experience set based on the match score.

4. The system of claim 2 , wherein the match score is determined based on at least one of a transaction time range, transaction merchant codes, transaction physical proximity, transaction temporal proximity, or a combination thereof.

5. The system of claim 1 , wherein each of the experience matchers is configured for execution by the one or more processors to perform a search and return operation in real time with a delay of less than ten minutes after receiving the list of transactions.

6. The system of claim 1 , wherein the transaction hub is configured for execution by the one or more processors to receive the transaction stream in real time with a delay of less than ten minutes after a transmittal of a new transaction to be appended to the list of transactions.

7. The system of claim 1 , wherein the system is configured to be modified by addition of new experience matchers.

8. The system of claim 7 , wherein the new experience matchers are created based on user input.

9. The system of claim 7 , wherein the new experience matchers are created by the machine learning model, the model including at least one of continuous learning, reinforcement learning, or retraining, wherein the machine learning model is configured to update an existing experience matcher to create a new experience matcher.

10. A computer-implemented method for forming an experience set from a transaction list from a financial services computer network, the method comprising:
obtaining by a transaction hub a past transaction list from the financial services computer network;
receiving by the transaction hub a transaction from the financial services computer network;
updating by the transaction hub the past transaction list into a current transaction list by adding in the received transaction;
sending by the transaction hub the current transaction list to each of a plurality of experience matchers;
analyzing by each of the experience matchers the current transaction list to identify in each of the experience matchers, by searching within the list of transactions, an experience set having an experience set type;
identifying a set of the transactions as being encompassed by the experience set based on one or more of a location or time associated with the transactions; and
receiving by the transaction hub a first identified experience set from a first of the experience matchers and a second identified experience set from a second of the experience matchers based on the list of transactions;
in response to the first identified experience set and the second identified experience set sharing at least one of the transactions, analyzing by a machine learning model the shared transaction, the first identified experience set, the second identified experience set, and historical transactions and experience sets of the user, the machine learning model being trained to prioritize allocation of transactions to experience sets on a per user basis by learning a hierarchy of experience sets;
choosing, by the machine learning model, one of the first identified experience set or the second identified experience set to be associated with the shared transaction;
assigning the shared transaction to the chosen experience set;
storing the chosen experience set by the transaction hub in a database;
providing transaction details associated with the chosen experience set for display to a user in a graphical user interface; and
continuously updating the machine learning model by retraining the machine learning model with the assignment of the shared transaction and the chosen experience set.

11. The method of claim 10 , wherein the current transaction list contains transactions from a predetermined time range.

12. The method of claim 11 , wherein the time range is thirty days.

13. The method of claim 10 , wherein the received transaction is received in real time with a delay of less than ten minutes.

14. The method of claim 10 , wherein the received first identified experience set is received in real time with a delay of less than ten minutes.

15. The method of claim 10 , further comprising receiving from each of the experience matchers a match score.

16. The method of claim 15 , wherein the match score is determined based on transaction time, transaction type, transaction distance, and transaction interval.

17. The method of claim 10 , further comprising revising the plurality of experience matchers.

18. The method of claim 10 , further comprising revising the first experience matcher to identify a different experience set.

19. The method of claim 18 , wherein the different experience set is created by the machine learning model, the model including at least one of continuous learning, reinforcement learning, or retraining, wherein

Assignees

• Capital One Services Llc

Inventors

• Hart Colin
• Ji Jason
• Colevas Alexandra A
• Quadros Steven
• Malik Adnan
• Alexander Jared M
• Golder Scott
• Abbott Allison
• Horton Matthew
• SAGAN Emma
• Emmott Kevan
• Berglund Christine Ann

Classifications

• G06Q40/02 Primary

  Banking, e.g. interest calculation or account maintenance (credit or loans G06Q40/03) · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F16/23

  Updating · CPC title