Systems and methods to quantify consumer sentiment based on transaction data

• US 9,384,493 B2
• Filed: 02/28/2013
• Issued: 07/05/2016
• Est. Priority Date: 03/01/2012
• Status: Active Grant

• Current Assignee: Visa International Service Association (Visa, Inc.)
• Original Assignee: Visa International Service Association (Visa, Inc.)
• Inventors: Harris, Theodore, Faith, Patrick, Siegel, Kevin Paul, Edington, Scott
• Primary Examiner(s): Singh, Gurkanwaljit
• Application Number:
  US13/780,409
• Publication Number:
  US 20130231974A1
• Time in Patent Office: 1,223 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

A computing apparatus is configured to quantify consumer sentiment at an aggregated or micro level using transaction data that records the transactions processed by a transaction handler of a payment system. A quantification model is generated based on correlating transaction data with respective emotional content indices extracted from data sources, such as regional news, weather, stock markets, movie themes, local sports, employment, traffic conditions, etc. Using the quantification model, consumer sentiment can be evaluated at various granularity levels, based on the granularity of the user group and the time period of the transaction data used in the quantification model.

Claims

1. A computer-implemented method, comprising:

providing a computing apparatus having

a transaction handler configured in an electronic payment processing network connecting separate computers, including

first computers controlling consumer accounts from which payments of transactions are made in the electronic payment processing network;

second computers controlling merchant accounts in which the payments are received via the electronic payment processing network; and

transaction terminals configured to initiate the transactions in the electronic payment processing network using identifications of the consumer accounts;

a data warehouse coupled with the transaction handler and storing;

first transaction data recording payment transactions processed by the transaction handler in a first period of time, and

second transaction data recording payment transactions processed by the transaction handler in a second period of time; and

a portal coupled with the data warehouse;

receiving, in the portal of the computing apparatus using a communication channel outside the electronic payment processing network, internet content indicative of consumer sentiment during the first period of time;

evaluating, by the computing apparatus, consumer sentiment values based on the internet content published during the first period of time;

correlating, by the computing apparatus using a machine learning technique, the first transaction data recording the payment transactions processed by the transaction handler in the first period of time with the consumer sentiment values evaluated based on the internet content published during the first period of time;

generating, by the computing apparatus, a computer quantification model of consumer sentiment from the correlating;

training the computer quantification model based on adjusting parameters of the quantification model to reduce differences between

the consumer sentiment values evaluated based on internet content published during the first period of time, and

numerical values computed from applying the first transaction data in the first period of time to the quantification model;

adjusting regional and temporal differences in emotional sentiment for the numerical values in the training of the quantification model;

receiving, by the computing apparatus from the data warehouse, the second transaction data recording payment transactions of a group of users during the second period of time;

applying, by the computing apparatus, the second transaction data to the quantification model;

determining, by the computing apparatus through the applying of the second transaction data to the computer quantification model, a numerical value of consumer sentiment of the group of the users during the second period of time;

identifying, by the computing apparatus, offers based at least in part on the numerical value of consumer sentiment of the group of the users; and

communicating, by the portal using a channel outside the electronic payment processing network, the offers to the group of the users.

2. The method of claim 1, wherein the group of users are identified based on a geographic region.

3. The method of claim 2, further comprising:

targeting offers based at least in part on the numerical value indicative of consumer sentiment of the group of the users.

4. The method of claim 1, wherein the internet content includes at least one of:

news feeds, social media feeds, and market reports.

5. The method of claim 1, the correlating further comprising:

training the quantification model based on adjusting parameters of the quantification model to reduce differences between

the consumer sentiment values evaluated based on internet content published during the first period of time, and

numerical values computed from applying transaction data of the first period of time to the quantification model.

6. The method of claim 5, further comprising:

adjusting regional and temporal differences in emotional sentiment for the first numerical values to train the quantification model.

7. The method of claim 1, further comprising:

collecting content indicative of consumer sentiment;

determining emotional content indices based on the content; and

correlating the emotional content indices with transaction data to establish the quantification model.

8. The method of claim 7, wherein the content includes at least one of:

regional news, weather, stock markets, movie themes, local sports, employment data, and traffic conditions.

9. The method of claim 7, wherein the correlating is based on a machine learning technique.

10. The method of claim 7, wherein the emotional content indices are associated with groups of users and periods of times;

and the correlating includes adjusting the quantification model to match quantification results with the emotional content indices, wherein the quantification results are obtained via applying transaction data of the groups of users in the periods of times to the quantification model.

11. The method of claim 1, further comprising:

monitoring consumer sentiment at a granularity that is based on a size of the group of the users and a length of the time period.

12. The method of claim 11, further comprising:

adjusting the granularity via the size of the group of the users and the length of the time period of the payment transactions based on which the numerical value is computed using the quantification model.

13. A tangible, non-transitory computer-storage medium storing instructions configured to instruct a computing apparatus to perform a method, the method comprising:

receiving, in a portal of the computing apparatus using a communication channel outside an electronic payment processing network, internet content indicative of consumer sentiment during a first period of time, wherein the computing apparatus including

a transaction handler configured in the electronic payment processing network connecting separate computers, including

first computers controlling consumer accounts from which payments of transactions are made in the electronic payment processing network;

second computers controlling merchant accounts in which the payments are received via the electronic payment processing network; and

transaction terminals configured to initiate the transactions in the electronic payment processing network using identifications of the consumer accounts;

a data warehouse coupled with the transaction handler and storing;

first transaction data recording payment transactions processed by the transaction handler in a first period of time, and

second transaction data recording payment transactions processed by the transaction handler in a second period of time; and

a portal coupled with the data warehouse;

evaluating, by the computing apparatus, consumer sentiment values based on the internet content published during the first period of time;

correlating, by the computing apparatus using a machine learning technique, the first transaction data recording the payment transactions processed by the transaction handler in the first period of time with the consumer sentiment values evaluated based on the internet content published during the first period of time;

generating, by the computing apparatus, a computer quantification model of consumer sentiment from the correlating;

training the computer quantification model based on adjusting parameters of the quantification model to reduce differences between

the consumer sentiment values evaluated based on internet content published during the first period of time, and

numerical values computed from applying the first transaction data in the first period of time to the quantification model;

adjusting regional and temporal differences in emotional sentiment for the numerical values in the training of the quantification model;

receiving, by the computing apparatus from the data warehouse, the second transaction data recording payment transactions of a group of users during the second period of time;

applying, by the computing apparatus, the second transaction data to the quantification model;

determining, by the computing apparatus through the applying of the second transaction data to the computer quantification model, a numerical value of consumer sentiment of the group of the users during the second period of time;

identifying, by the computing apparatus, offers based at least in part on the numerical value of consumer sentiment of the group of the users; and

communicating, by the portal using a channel outside the electronic payment processing network, the offers to the group of the users.

14. A computing apparatus having at least one processor and a memory storing instructions configured to instruct the at least one processor to perform operations, the computing apparatus comprising:

a transaction handler configured in an electronic payment processing network connecting separate computers, including

first computers controlling consumer accounts from which payments of transactions are made in the electronic payment processing network;

second computers controlling merchant accounts in which the payments are received via the electronic payment processing network; and

transaction terminals configured to initiate the transactions in the electronic payment processing network using identifications of the consumer accounts;

a data warehouse coupled with the transaction handler and configured to store

first transaction data recording payment transactions processed by the transaction handler in a first period of time, and

second transaction data recording payment transactions of a group of users during a second period of time;

a portal configured to receive, using a communication channel outside the electronic payment processing network, internet content indicative of consumer sentiment during the first period of time, wherein the computing apparatus is configured to evaluate consumer sentiment values based on the internet content published during the first period of time;

a learning engine coupled with the data warehouse and configured to

perform correlation, using a machine learning technique, between the payment transactions processed by the transaction handler in the first period of time and the consumer sentiment values evaluated based on the internet content published during the first period of time,

generate a computer quantification model of consumer sentiment from the correlation,

train the computer quantification model based on adjusting parameters of the quantification model to reduce differences between

the consumer sentiment values evaluated based on internet content published during the first period of time, and

numerical values computed from applying the first transaction data in the first period of time to the quantification model;

adjust regional and temporal differences in emotional sentiment for the numerical values in the training of the quantification model; and

a rule engine coupled with the data warehouse and configured to

apply the second transaction data to the quantification model,

determine, through applying the second transaction data to the computer quantification model, a numerical value of consumer sentiment of the group of the users during the second period of time; and

identifying, by the computing apparatus, offers based at least in part on the numerical value of consumer sentiment of the group of the users,

wherein the portal is further configured to communicate, using a channel outside the electronic payment processing network, the offers to the group of the users.

15. The computing apparatus of claim 14, wherein the portal is further configured to provide access to numerical values computed based on the quantification model and indicative of consumer sentiment of groups of the users and receive model performance feedback for the numerical values computed based on the quantification model and accessed via the portal;

and the computing apparatus further comprises;

an analytics engine configured to analyze the model performance feedback to provide input data for the learning engine to adjust the quantification model.

16. The computing apparatus of claim 15, where granularity of the numerical values computed based on the quantification model and indicative of consumer sentiment of the groups of the users is based on sizes of the groups of the users.