Behavioral data mining processes for generating pairwise item comparisons

• US 8,234,183 B2
• Filed: 01/20/2012
• Issued: 07/31/2012
• Est. Priority Date: 06/05/2008
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Smith, Brent R.
• Primary Examiner(s): Smith, Jeffrey A
• Assistant Examiner(s): Desai, Resha
• Application Number:
  US13/355,388
• Publication Number:
  US 20120158552A1
• Time in Patent Office: 193 Days
• Field of Search: 705 261- 272, 705 141- 1473
• US Class Current: 705/26.64
• CPC Class Codes: G06Q 30/0629 for generating comparisons
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0641 Shopping interfaces
• CPC Class Codes: G06Q 30/0643 Graphical representation of items or shoppers

Abstract

Data mining systems and methods are disclosed for generating data that is helpful to users in selecting between items represented in an electronic data repository, such as an electronic catalog. One disclosed data mining method generates pairwise comparison data for particular pairs of items. The pairwise comparison data for a given item pair reveals a tendency of users who consider both items in the pair to select one item over the other. The pairwise comparison data may be appropriately exposed to users of the electronic repository. For instance, an item detail page for item A may be supplemented with an indication that users who view both item A and item B select item B a specified percentage of the time. Another data mining method uses item viewing histories and item purchase histories of users in combination to identify pairs of items that are good candidates for being recommended in combination.

Claims

1. A computer-implemented method of assisting users in selecting between a first item and a second item represented in an electronic data repository, the method comprising:

recording, in computer storage, user-specific event sequences reflective of actions performed by users in connection with items represented in the electronic data repository, said actions including item consideration actions and item selection actions performed on user computing devices;

identifying, based on a computerized analysis of the user-specific event sequences, a plurality of user-specific event sequences in which a corresponding user (1) considered both the first item and the second item, and (2) selected either the first item or the second item;

generating, by a computer system, based on the identified plurality of user-specific event sequences, pairwise comparison data representing a tendency of users who consider both the first item and the second item to select one such item over the other; and

outputting a visual representation of the pairwise comparison data for display to users in conjunction with descriptions of the first and second items.

2. The method of claim 1, wherein the items are catalog items represented in an electronic catalog, and wherein identifying the plurality of user-specific event sequences comprises treating a user'"'"'s visit to an item detail page of the electronic catalog as consideration by the user of an item featured on the item detail page.

3. The method of claim 2, wherein the item selection actions are item acquisition actions.

4. The method of claim 1, wherein the items are catalog items represented in an electronic catalog, and the pairwise comparison data represents a tendency of users who view both the first item and the second item to select one item over the other for purchase.

5. The method of claim 1, wherein the item consideration actions are based on viewing events, and the item selection actions are based on item acquisitions events.

6. The method of claim 1, wherein the visual representation of the pairwise comparison data comprises a first percentage value corresponding to the first item and a second percentage value corresponding to the second item, such that the sum of the first and second percentage values is 100.

7. A data mining system, comprising:

a computer data repository that stores records of user-specific event sequences, said user-specific event sequences reflective of actions performed by users via user computing devices in connection with items represented in an electronic data repository, said actions including item consideration actions and item selection actions; and

a computer system comprising at least one physical computer, said computer system programmed to at least;

identify, based on an analysis of the user-specific event sequences, a plurality of user-specific event sequences in which a corresponding user (1) considered both a first item and a second item, and (2) selected either the first item or the second item, but not both;

generate, based on the identified plurality of user-specific event sequences, pairwise comparison data reflective of a tendency of users who consider both the first item and the second item to select one such item over the other; and

output a representation of the pairwise comparison data for display to users.

8. The data mining system of claim 7, wherein the items are catalog items represented in an electronic catalog, and the computer system, in generating the pairwise comparison data, is programmed to treat a user'"'"'s visit to an item detail page of the electronic catalog as consideration by the user of an item represented on the item detail page.

9. The data mining system of claim 8, wherein the item selection actions are item acquisition actions.

10. The data mining system of claim 7, wherein the items are catalog items represented in an electronic catalog, and the pairwise comparison data reveals a tendency of users who view both the first item and the second item to select one item over the other for purchase.

11. The data mining system of claim 7, wherein the item consideration actions are viewing actions, and the item selection actions are item purchase actions.

12. The data mining system of claim 7, wherein the representation of the pairwise comparison data comprises a first percentage value corresponding to the first item and a second percentage value corresponding to the second item, such that the sum of the first and second percentage values is 100.

13. Non-transitory computer storage that stores executable program code that directs a computer system to perform a process that comprises:

storing records of user-specific event sequences, said user-specific event sequences reflective of actions performed by users in connection with items represented in an electronic data repository, said actions including item consideration actions and item selection actions;

identifying, based on an analysis of the user-specific event sequences, a plurality of user-specific event sequences in which a corresponding user (1) considered both a first item and a second item, and (2) selected either the first item or the second item, but not both;

generating, based on the identified plurality of user-specific event sequences, pairwise comparison data reflective of a tendency of users who consider both the first item and the second item to select one such item over the other; and

outputting a representation of the pairwise comparison data for display to users in conjunction with identifiers of the first and second items.

14. The non-transitory computer storage of claim 13, wherein the items are catalog items represented in an electronic catalog, and the pairwise comparison data represents a tendency of users who view both the first item and the second item to purchase one item over the other.

15. The non-transitory computer storage of claim 13, wherein the item selection actions are item acquisition actions.

16. The non-transitory computer storage of claim 13, wherein the representation of the pairwise comparison data comprises a first percentage value representing a percentage of users who selected the first item, and a second percentage value representing a percentage of users who selected the second item, such that the sum of the first and second percentage values is 100.

17. A computer-implemented method of assisting users in selecting between a first item and a second item represented in an electronic catalog, the method comprising:

receiving, from a user computing device, a request for an item detail page for a first catalog item;

retrieving pairwise comparison data representing a tendency of users who view both the first catalog item and a second catalog item to select the first catalog item over the second catalog item, said pairwise comparison data based on a plurality of recorded event sequences in which a corresponding user (1) viewed both the first and second catalog items and (2) purchased either the first catalog item or the second catalog item, but not both;

incorporating a visual representation of the pairwise comparison data into the item detail page; and

sending to the user computing device, in response to the request, the item detail page with the incorporated visual representation of the pairwise comparison data;

said method performed by a computer system that comprises a physical server.

18. The method of claim 17, wherein the visual representation of the pairwise comparison data comprises a first percentage value representing a percentage of users who purchased the first item, and a second percentage value representing a percentage of users who purchased the second item, such that the sum of the first and second percentage values is 100.