Automated selection of three of more items to recommend as a bundle

• US 8,438,052 B1
• Filed: 04/20/2009
• Issued: 05/07/2013
• Est. Priority Date: 04/20/2009
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Chanda, Gaurav, Smith, Brent R, Whitman, Ronald M
• Primary Examiner(s): Glass, Shay S
• Application Number:
  US12/426,441
• Time in Patent Office: 1,478 Days
• Field of Search: 705/14
• US Class Current: 705/7.11
• CPC Class Codes: G06Q 30/0207 Discounts or incentives, e.g. coupons or rebates
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products

Abstract

Automated processes are disclosed for identifying bundles of three or more items to suggest to users. In one embodiment, item action data, such as item purchase data reflective of the item purchases of a user population, is used to detect pairwise (item to item) associations between particular items represented in an electronic catalog. These pairwise associations are then used (optionally in combination with other information) to select bundles of three or more items to recommend to users. The bundles may, for example, be suggested on item detail pages of the electronic catalog.

Claims

1. A computer-implemented method of selecting items to include in item bundles, the method comprising:

maintaining purchase data in computer storage, said purchase data reflective of item purchases of each of a plurality of users;

detecting pairwise purchase relationships between particular items based, at least in part, on how frequently particular items are purchased in combination by users, as determined by analyzing said purchase data;

identifying, based on the detected pairwise purchase relationships, a set of three pairwise purchase relationships said set composed of a pairwise purchase relationship between a first item and a second item, a pairwise purchase relationship between the first item and a third item, and a pairwise purchase relationship between the second item and the third item, said first, second and third items each having a respective item detail page in an electronic catalog; and

based on at least the identified set of three pairwise purchase relationships, supplementing the item detail page of the first item with (1) a description of a three-item bundle composed of the first, second and third items, and (2) an option to select the three-item bundle for purchase;

wherein the method, including maintaining the purchase data, detecting the pairwise purchase relationships, identifying the set of three pairwise purchase relationships, and supplementing the item detail page, is performed programmatically by a computerized system that comprises one or more physical computing devices.

2. The method of claim 1, wherein the method comprises programmatically inferring that the first, second, and third items are suitable for inclusion in the bundle based on said pairwise purchase relationships between the first and second items, first and third items, and second and third items.

3. The method of claim 1, wherein identifying the set of three pairwise purchase relationships comprises:

identifying a first plurality of items that each have a pairwise purchase relationship with the first item;

identifying a second plurality of items that each have a pairwise purchase relationship with the second item; and

determining that said third item is included in both said first plurality of items and said second plurality of items.

4. The method of claim 1, wherein the step of supplementing the item detail page of the first item is performed without regard to how frequently the first, second, and third items are actually purchased in combination.

5. The method of claim 1, wherein the step of supplementing the item detail page is performed based further on a determination that attributes of the first, second and third items do not satisfy a set of bundle exclusion criteria.

6. The method of claim 1, further comprising collecting feedback data regarding user responses to recommendations of said three-item bundle, and taking said feedback data into consideration to determining whether to continue to recommend said three-item bundle.

7. The method of claim 1, wherein the option to select the three-item bundle for purchase comprises a control that is selectable by a user to add the first, second and third items collectively to an electronic shopping cart.

8. The method of claim 7, further comprising adding the first, second and third items to an electronic shopping cart in response to user selection of said control.

9. A system for selecting items to include in bundles, the system comprising:

a data storage system that stores purchase data reflective of item purchases of each of a plurality of users; and

a computer system comprising one or more physical computing devices, said computer system programmed to implement a process that comprises;

detecting pairwise purchase relationships between particular items based, at least in part, on how frequently particular items are purchased in combination by users, as determined by analyzing said purchase data;

identifying, based on the detected pairwise purchase relationships, a set of three pairwise purchase relationships, said set composed of a pairwise purchase relationship between a first item and a second item, a pairwise purchase relationship between the first item and a third item, and a pairwise purchase relationship between the second item and the third item, said first, second and third items each having a respective item page in an electronic catalog; and

based at least partly on the identified set of three pairwise purchase relationships, causing the item page of the first item to be supplemented with (1) a description of a three-item bundle composed of the first, second and third items, and (2) an option to select the three-item bundle for purchase.

10. The system of claim 9, wherein the computer system is programmed to cause the item page to be supplemented with said description and option without regard to how frequently the first, second and third items are actually purchased in combination.

11. The system of claim 9, wherein the computer system is programmed to identify the set of three pairwise purchase relationships by a process that comprises:

identifying a first plurality of items that each have a pairwise purchase relationship with the first item;

identifying a second plurality of items that each have a pairwise purchase relationship with the second item; and

determining that said third item is included in both said first plurality of items and said second plurality of items.

12. The system of claim 9, wherein the computer system is additionally programmed to analyze attributes of the first, second and third items to assess whether the first, second and third items are suitable for inclusion in the three-item bundle.

13. The system of claim 12, wherein said attributes include respective prices of the first, second and third items.

14. The system of claim 12, wherein said attributes include respective categorizations of the first, second and third items.

15. The system of claim 9, wherein the computer system is additionally programmed to use implicit user feedback on the three-item bundle to assess whether to continue causing said item page to be supplemented with said description and option.

16. The system of claim 9, wherein the option to select the three-item bundle for purchase comprises a control that is selectable by a user to add the first, second and third items collectively to an electronic shopping cart.

17. The system of claim 9, wherein the computer system is responsive to user selection of said control by adding the first, second and third items to the electronic shopping cart.