Extrapolation-based creation of associations between search queries and items

• US 8,090,625 B2
• Filed: 04/22/2011
• Issued: 01/03/2012
• Est. Priority Date: 06/16/2006
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Yi, Jin Y.
• Primary Examiner(s): SHAAWAT, MUSSA A
• Application Number:
  US13/092,439
• Publication Number:
  US 20110196895A1
• Time in Patent Office: 256 Days
• Field of Search: None
• US Class Current: 705/26.7
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 30/0629 for generating comparisons
• CPC Class Codes: G06Q 30/0631 Item recommendations

Abstract

Behavior-based associations, such as item-to-item or query-to-item associations, are extrapolated to other items to create new associations. The items to which the associations are extrapolated may be “behavior deficient” items, or items for which the quantity of collected user activity data is insufficient to create meaningful or reliable behavior-based associations. The behavior-based associations are extrapolated based on content-based associations, or another type of “substitutability” association, between items. The items can be any type of item (e.g., products, web sites, documents, etc.) for which user behaviors (e.g., purchases, accesses, downloads, etc.) can be monitored and analyzed to detect behavior-based associations, and for which item content or other available information can be used to assess item substitutability.

Claims

1. A computer-implemented method of creating associations between search queries and items, the method comprising:

generating a first value that represents a strength of a behavior-based association between a search query and a first item, said behavior-based association based at least partly on item selection actions of users who have submitted the search query, said behavior-based association reflecting a tendency of users who submit the search query to select the first item;

generating a second value that represents a strength of a substitutability association between the first item and a second item, said substitutability association based at least partly on an analysis of content of the first and second items; and

creating, in computer storage, a mapping that associates the search query with the second item, said mapping created based at least partly on both (1) the detected behavior-based association between the search query and the first item, and (2) the detected substitutability association between the first item and the second item, wherein creating said mapping comprises generating, based at least partly on the first and second values, a third value that represents a strength of an association between the search query and the second item;

said method performed in its entirety by a computer system that comprises one or more computers.

2. The method of claim 1, wherein the method comprises creating said mapping based further on a determination that the second item is a behavior-deficient item.

3. The method of claim 2, further comprising:

using said mapping as a basis for causing the second item to be presented to users; and

phasing out or terminating such usage of the mapping in response to the second item becoming a non-behavior-deficient item.

4. The method of claim 1, further comprising using the mapping to cause the second item to be included in search results presented to a user who submits the search query.

5. The method of claim 1, further comprising using the mapping to rank search results for display.

6. The method of claim 1, wherein the mapping is based additionally on (3) a detected behavior-based association between the search query and a third item, and (4) a detected substitutability association between the second item and the third item.

7. The method of claim 1, wherein the third value is additionally dependent on:

a fourth value representing a strength of a behavior-based association between the search query and a third item, and

a fifth value representing a strength of a substitutability association between the second item and the third item.

8. The method of claim 1, wherein the first and second items are respective web pages.

9. The method of claim 1, wherein the first and second items are respective products represented in an electronic catalog.

10. Non-transitory computer storage having stored thereon executable code that directs a computer system to perform a process that comprises:

generating a first value that represents a strength of a behavior-based association between a search query and a first item, said behavior-based association based at least partly on item selection actions of users who have submitted said search query;

generating a second value that represents a strength of a substitutability association between the first item and a second item; and

creating, in computer storage, an association between the search query and the second item, said association created based at least partly on both (1) the detected behavior-based association between the search query and the first item, and (2) the detected substitutability association between the first item and the second item, wherein creating said association comprises generating, based at least partly on the first and second values, a third value that represents a strength of the association between the search query and the second item.

11. The non-transitory computer storage of claim 10, wherein the process comprises creating said association between the search query and the second item based further on a determination that the second item is a behavior-deficient item.

12. The non-transitory computer storage of claim 11, wherein the process further comprises:

using the created association as a basis for causing the second item to be presented to users; and

phasing out or terminating such usage of the created association in response to the second item becoming a non-behavior-deficient item.

13. The non-transitory computer storage of claim 10, further comprising using the created association to cause the second item to be included in search results presented to a user.

14. The non-transitory computer storage of claim 10, further comprising using the created association to rank search results for display.

15. The non-transitory computer storage of claim 10, wherein the association between the search query and the second item is created based additionally on (3) a detected behavior-based association between the search query and a third item, and (4) a detected substitutability association between the second item and the third item.

16. The non-transitory computer storage of claim 10, wherein the third value is additionally dependent on:

a fourth value representing a strength of a behavior-based association between the search query and a third item, and

a fifth value representing a strength of a substitutability association between the second item and the third item.

17. The non-transitory computer storage of claim 10, wherein the first and second items are respective web pages.

18. The non-transitory computer storage of claim 17, wherein the process comprises detecting said substitutability association based at least partly on a comparison of content of said web pages.

19. The non-transitory computer storage of claim 10, wherein the first and second items are respective products represented in an electronic catalog.

20. The non-transitory computer storage of claim 19, wherein the process comprises detecting said substitutability association based at least partly on a comparison of attributes of said products.

21. The non-transitory computer storage of claim 10, in combination with said computer system, wherein the computer system is programmed with said executable code to perform said process.

22. A system for detecting associations between search queries and items:

a computer system comprising one or more computers, said computer system configured to implement;

a substitutability association mining component that detects substitutability associations between particular items based, at least in part, on item content, said substitutability association mining component configured to generate a first value representing a strength of a substitutability association between a first item and a second item;

a behavioral association mining component that detects behavioral associations between particular search queries and items based, at least in part, on search query submission actions and item selection actions of a plurality of users, said behavioral association mining component configured to generate a second value representing a strength of a behavioral association between a search query and the first item; and

an extrapolation component that creates additional associations between search queries and items by combining the detected substitutability associations with the detected behavioral associations, said extrapolation component configured to extrapolate a behavioral association between the search query and the first item to a second item to create an association between the search query and the second item, said extrapolation component configured to generate a third value representing a strength of the association between the search query and the second item, said third value being dependent upon at least the first value and the second value.

23. The system of claim 22, further comprising a search engine that uses the additional associations created by the extrapolation component to rank search results for display to users.

24. The system of claim 22, wherein the third value is additionally dependent on:

a fourth value representing a strength of a behavioral association between the search query and a third item, and

a fifth value representing a strength of a substitutability association between the second item and the third item.