Technique to aggregate merchant level information for use in a supervised learning model to detect recurring trends in consumer transactions

Patent metadata

Field	Value
Publication number	US-11144935-B2
Application number	US-201916657385-A
Country	US
Kind code	B2
Filing date	Oct 18, 2019
Priority date	Oct 18, 2019
Publication date	Oct 12, 2021
Grant date	Oct 12, 2021

Abstract

A method is disclosed, comprising: aggregating a plurality of sets of transactions, each set of transactions comprising transactions related to an account-merchant pairing; determining variables characterizing a recurrence period based on transaction dates of the transactions in the each set; predicting the recurrence period for a transaction related to the account-merchant pairing for each customer of the plurality of customers; aggregating another set of transactions between the plurality of customers and the merchant; evaluating a distribution of the recurrence period for each customer within range of a distant point; and based on the evaluation of the distribution of the recurrence period for each customer, generating a probability of the merchant having a recurrent transaction with the customer. The account-merchant pairing may comprise a customer account and a merchant of a plurality of customers and merchants. Another set of transactions may comprise transactions in the plurality of sets of transactions.

First claim

What is claimed is:

1. A method, comprising:
storing, by an application server, information gathered from a plurality of data sources in a database, the information comprising merchant identification information of a plurality of merchants and a plurality of transactions of a plurality of customer accounts is gathered for categorizing a transaction recurrency between a merchant of the plurality of merchants and customer account of the plurality of customer accounts;
aggregating, by the application server, a plurality of sets of transactions based on the gathered information stored in the database, each set of transactions is related to an account-merchant pairing, wherein the account-merchant pairing includes a customer account the plurality of customer accounts and a merchant of the plurality of merchants;
splitting each set of the transactions into a plurality of first subset of transactions and corresponding second subset of transactions for generating expanded test data sets, wherein transactions are split based on a split date;
training a machine learning algorithm by using target label variables, the target label variables are generated based on the second subset of transactions according to predicted transaction dates in the second subset of transactions of the corresponding first subset of transactions of the expanded test data sets and predetermined matching criteria for training a machine learning algorithm, wherein the target label variables identify recurrency of transactions in the first subset of transactions, wherein the matching criteria comprises a number of predictions and a date tolerance, and wherein the predicted transaction dates in the second subset of transactions within the date tolerance are at least the number of predictions in the matching criteria;
converting, by the application server, transactions dates of the transactions in each first subset of transactions into ordinal dates;
mapping, by the application server, the ordinal dates into a phase space, wherein the phase space represents a billing cycle;
determining, by the application server, a mean coordinate based on the mapped ordinal dates;
determining, by the application server, a vector strength, a coverage, and a redundancy, wherein the vector strength, the coverage, and the redundancy are phase variables characterizing a recurrence period;
predicting, by the application server, the recurrence period for a transaction related to the account-merchant pairing for each customer of the plurality of customers;
aggregating, by the application server, another set of transactions between the plurality of customers and the merchant, wherein the another set of transactions comprises transactions in the plurality of sets of transactions;
evaluating, by the application server, a distribution of the recurrence period for each customer based on a plurality of multi-dimensional distribution points representing the recurrence period within a range of a point representing predetermined values of the vector strength, the coverage, and the redundancy; and
based on the evaluation of the distribution of the recurrence period for each customer, generating, by the application server, a probability of the merchant having a recurrent transaction with the customer account using the trained machine learning algorithm based on the generated target label variables.

2. The method of claim 1 , wherein the first subset of transactions are used as a training subset and determined based on occurring prior to the split date and the second subset of transactions are used as a testing subset and determined based on occurring after the split date.

3. The method of claim 1 , further comprising:
determining a standard deviation of difference between the transaction dates of the transactions in each first subset of transactions;
determining an average difference between the transaction dates of the transactions in each first subset of transactions; and
determining a ratio between the standard deviation of difference between the transaction dates and the average difference between the transaction dates,
wherein the standard deviation of the transaction dates, the average difference between the transaction dates, and the ratio of the standard deviation of the transaction dates to the average difference between the transaction dates are delta time variables characterizing the recurrence period.

4. The method of claim 1 ,
wherein the vector strength is a distance between a (0,0) coordinate of a unit circle and the mean coordinate,
wherein the coverage is a count of billing cycles in the phase space having at least one transaction, and
wherein the redundancy is a percentage of billing cycles having more than one transaction.

5. The method of claim 1 , further comprising determining a modified vector strength as

r
adjusted

=

1
-

2
π

⁢

arccos
⁡

(
r
)

,

wherein r adjusted is the modified vector strength and r is the vector strength.

6. The method of claim 1 , further comprising determining a probability of the transactions in each first subset of transactions for the recurrence period based on a Euclidean distance between the point representing the predetermined values of the vector strength, the coverage, and the redundancy and another point representing another value of the vector strength, the coverage, and the redundancy as (the vector strength, the coverage, the redundancy) for phase spaces, wherein each distance between phase variable coordinates corresponds to one of a weekly, a biweekly, a monthly, a bimonthly, a quarterly, a semi-annually, and a yearly period.

7. The method of claim 1 , wherein aggregating the plurality of sets of transactions with each set of transactions comprising transactions related to the account-merchant pairing comprises:
ordering the transactions in a chronological order; and
discarding transactions at a beginning and at an end of the chronologically ordered transactions within a user specified threshold value.

8. The method of claim 1 , wherein the machine learning algorithm is a supervised machine learning algorithm.

9. A system, comprising:
a memory for storing instructions; and
a processor, communicatively coupled to the memory, configured to execute the instructions, the instructions causing the processor to:
store information gathered from a plurality of data sources in a database, the information comprising merchant identification information of a plurality of merchants and a plurality of transactions of a plurality of customer accounts is gathered for categorizing a transaction recurrency between a merchant of the plurality of merchants and customer account of the plurality of customer accounts;
aggregate a plurality of sets of transactions based on the gathered information stored in the database, each set of transactions is related to an account-merchant pairing, wherein the account-merchant pairing includes a customer account of the plurality of customer accounts and a merchant of the plurality of merchants;
split each set of the transactions into a plurality of

Assignees

• Capital One Services Llc

Inventors

• Jumper Daniel
• GERSTLE Jeremy
• Anthony Olufemi
• Boroumand Jonathan
• Virbitsky Erik
• Walters Michael
• Bosede Aniver

Classifications

• G06Q30/0201 Primary

  Market modelling; Market analysis; Collecting market data · CPC title

• G06Q40/12

  Accounting · CPC title

• G06Q10/06315

  Needs-based resource requirements planning or analysis · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F16/248

  Presentation of query results · CPC title