Method and system for extracting customer attributes

• US 20070112733A1
• Filed: 11/14/2005
• Published: 05/17/2007
• Est. Priority Date: 11/14/2005
• Status: Abandoned Application

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Zhang, Alex, Beyer, Dirk, Shahoumian, Troy
• Application Number:
  US11/272,907
• Publication Number:
  US 20070112733A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A method for extracting customer attributes from a database includes providing information regarding one or more content, one or more metrics, customer ID and a reference date and extracting customer attributes based on the information.

Claims

1. A method for extracting customer attributes from a database, comprising:

providing information regarding content, a metric, customer identifier (ID) and a reference date; and

extracting customer attributes based on said information.

2. A method as defined in claim 1, wherein the customer attributes that are extracted are used to form a customer behavior model.

3. The method as defined in claim 2 further comprising using the customer behavior model to generate an advertisement, said advertisement comprising one selected from a group consisting of direct mail, web advertisement, and telephone solicitation.

4. A method as defined in claim 1, wherein the content comprises past offers or transactions that have been made to particular customers.

5. A method as defined in claim 1, wherein providing comprises providing a past purchase metric and past stimulus metrics.

6. A method as defined in claim 5, wherein providing information further comprises providing a stimulus response metric.

7. A method as defined in claim 5, wherein the past purchase metric comprises at least one from a group consisting of a stock keeping unit (SKU) and a value of interest.

8. A method as defined in claim 1, wherein the extracted attributes are extracted as a customer attribute table for each of a plurality of customer IDs.

9. A method as defined in claim 1 wherein extracting customer attributes comprises aggregating past customer transactions into said customer attributes grouped according to type of event, time interval and a metric to be aggregated.

10. A method as defined in claim 9, wherein the metric to be aggregated is selected from a group consisting of a stimulus response metric, a past purchase metric and a past stimulus metric.

11. A method as defined in claim 10, wherein the stimulus response metric is selected from a group consisting of revenue, profit, number of items ordered, number of categories ordered and an indicator.

12. A method as defined in claim 10, wherein the past purchase metric comprises transactions that are derived from past purchases made by customers.

13. A method as defined in claim 10, wherein the past stimulus metric comprises past marketing stimuli that have been presented to customers.

14. A method as defined in claim 10 wherein providing information also comprises providing content identification (ContentID) information and wherein extracting customer attributes comprises extracting customer attributes based also on the ContentID information, and the method further comprising generating a labeled dataset using the extracted customer attributes.

15. A method as defined in claim 14, further comprising using the labeled dataset as training data for a customer behavior model.

16. A method for extracting training datasets for customer behavior modeling from a database, comprising:

providing information describing an event, response metrics and a customer identifier (ID);

extracting customer attributes based on data associated with the event; and

labeling the extracted customer attributes based on the information.

17. A method as defined in claim 16 wherein extracting customer attributes is also based on a date.

18. A method as defined in claim 16, wherein the response metric(s) are selected from a group consisting of an indicator variable indicating whether the customer responded, a revenue of a transaction and a stated satisfaction of the customer with the product.

19. A method as defined in claim 16, wherein the event comprises a marketing stimulus.

20. A system, comprising:

means for providing information regarding one or more content, one or more metrics, customer ID and a reference date; and

means for extracting customer attributes based on said information.

21. A storage medium accessible to a computer, said storage medium containing software that, when executed by the computer, causes the computer to:

receive information regarding one or more content, one or more metrics, customer ID and a reference date; and

extract customer attributes based on said information.

22. The storage medium of claim 21 wherein the software further causes the computer form a customer behavior model using the extracted customer attributes.

23. The storage medium of claim 21 wherein the software further causes the processor to aggregate past customer transactions into said customer attributes grouped according to type of event, time interval and a metric to be aggregated

24. The storage medium of claim 21 wherein the software further causes the processor to generate a labeled dataset using the extracted customer attributes.

25. A method for extracting customer attributes from a database, comprising:

providing information regarding past transactions, a past purchase metric, a past stimulus metric, a stimulus response metric, customer identifier (ID) and a reference date;

extracting customer attributes based on said information, wherein extracting customer attributes comprises aggregating said past transactions into said customer attributes grouped according to type of event, time interval and a metric to be aggregated; and

forming a customer behavior model using the extracted customer attributes.