Automated selection of item categories for presenting item recommendations

• US 8,117,216 B1
• Filed: 08/26/2008
• Issued: 02/14/2012
• Est. Priority Date: 08/26/2008
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Chanda, Gaurav, Smith, Brent
• Primary Examiner(s): LE, UYEN T
• Application Number:
  US12/198,432
• Time in Patent Office: 1,267 Days
• Field of Search: None
• US Class Current: 707/754
• CPC Class Codes: G06F 16/9038 Presentation of query results

Abstract

A computer-implemented process for reducing category overlap obtains a set of source items and an initial set of item categories associated with the source items. An assessment of the overlap between different item categories in the initial set of item categories is used to selectively eliminate one or more item categories prior to presentation of the source items and associated categories to the user, thereby resulting in reduced overlap between the item categories presented. The resulting set of categories improves the ability of the user to efficiently and effectively access source items of interest, such as recommended items generated by a recommendation engine.

Claims

1. A computer-implemented method of facilitating the navigation of recommended items, the method comprising:

receiving, from a recommendation engine, an indication of a plurality of items selected by the recommendation engine to recommend to a user, said plurality of items selected by the recommendation engine based on information reflecting preferences of the user, each of the plurality of items being pre-assigned to at least one item category of a hierarchical browse structure;

determining, based on said pre-assigned categorizations, an initial set of item categories associated with the plurality of items, each of the item categories in the initial set being associated with one or more of the plurality of items;

selectively filtering out item categories from the initial set of item categories based at least in part on the degree to which particular item categories in the set overlap with other item categories in the set, to thereby create a subset of item categories to present to the user; and

providing, based on the created subset of item categories, a recommendations user interface that includes functionality for the user to navigate at least a portion of the plurality of items using the subset of item categories;

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

2. The method of claim 1, wherein the item categories are selectively filtered out of the initial set based additionally on a ranking of the plurality of items.

3. The method of claim 1, wherein the degree to which particular item categories in the set overlap with other item categories in the set is determined by calculating overlap scores for particular item categories, wherein the overlap score for a particular item category depends on an extent to which the particular item category shares items with one or more other item categories in said set.

4. The method of claim 3, wherein selectively filtering out item categories from the initial set of item categories comprises iteratively removing from the set the item category having the largest overlap score.

5. The method of claim 4, further comprising ceasing the iterative removal of item categories when no item categories remain having an overlap score exceeding a threshold.

6. The method of claim 3, wherein the overlap score for a first item category is determined by summing incremental overlap score contributions of each of a plurality of other item categories that overlap with the first item category.

7. The method of claim 1, wherein selectively filtering out item categories comprises giving different non-zero amounts of weight to some of the items selected by the recommendation engine than to other items selected by the recommendation engine.

8. The method of claim 1, wherein the method is performed in its entirety in real time in response to a request from the user for item recommendations.

9. The method of claim 1, wherein the method comprises calculating said degree to which particular item categories in the initial set overlap based on the items selected by the recommendation engine, without regard to additional items falling within the initial set of item categories.

10. The method of claim 9, wherein providing a recommendations user interface comprises providing a text cloud interface in which each item category of the subset is selectable by the user to view a corresponding subset of said plurality of items.

11. A computer system, comprising:

a recommendation engine configured to select a plurality of items to recommend to a user based on information regarding past behaviors of the user;

a category filtering system in communication with the recommendation engine, the category filtering system configured to determine an initial set of item categories associated with the plurality of items selected by the recommendation engine, and to selectively filter out item categories from the initial set of item categories based at least in part on the degree to which particular item categories in the set overlap with other item categories in the set, to thereby create a subset of item categories to present to the user; and

a server that provides a recommendations user interface that includes functionality for using the subset of item categories to interactively browse at least a portion of the plurality of items, said recommendations user interface omitting the filtered-out item categories.

12. The computer system of claim 11, wherein the category filtering system is configured to determine the degree to which particular item categories in the set overlap with other item categories in the set by calculating overlap scores for particular item categories, wherein the overlap score for a particular item category depends on an extent to which the particular item category shares items with one or more other item categories in said set.

13. The computer system of claim 12, wherein the category filtering system is configured to selectively filter out particular item categories from the initial set of item categories by a process that comprises iteratively removing from the set an item category having the largest overlap score.

14. The computer system of claim 12, wherein the category filtering system is configured to calculate an overlap score for a first item category by a process that comprises summing incremental overlap score contributions of each of a plurality of other item categories that overlap with the first item category.

15. The computer system of claim 11, wherein the category filtering system is configured to determine which item categories to filter out based in part on a ranking of the plurality of items, said ranking being specific to said user.

16. The computer system of claim 11, wherein the category filtering system is configured to give different non-zero amounts of weight to some of the items than to other items in determining which item categories to filter out.

17. The computer system of claim 11, wherein the recommendations user interface is a text cloud interface in which each item category in said subset is selectable to filter out from display items falling outside that item category.

18. Non-transitory computer storage comprising executable code that directs a computer system to perform a process that comprises:

receiving, from an items-of-interest engine, a representation of a set of source items identified as being of interest to a target user;

determining a set of item categories associated with said set of source items; and

selecting, from the set of item categories, a subset of item categories to use to present the set of source items to the target user, wherein selecting the subset of item categories comprises filtering out at least a first item category from the set based, at least in part, on a calculated measure of overlap between the first item category and one or more other item categories in said set.

19. The non-transitory computer storage of claim 18, wherein the process comprises calculating said measure of overlap based on categorizations of said source items.

20. The non-transitory computer storage of claim 19, wherein the process comprises giving different amounts of weight to different source items in calculating said measure of overlap.

21. The non-transitory computer storage of claim 18, wherein the subset of item categories is personalized for the target user.

22. The non-transitory computer storage of claim 18, wherein the process further comprises generating a user interface that includes functionality for the user to interactively browse the set of source items using the selected subset of item categories.

23. The non-transitory computer storage of claim 22, wherein the user interface is a text cloud interface in which each item category in said subset is selectable to view the source items falling therein.

24. The non-transitory computer storage of claim 18, in combination with said computer system, wherein the computer system is programmed with said executable code to perform said process.

25. The non-transitory computer storage of claim 18, wherein the process comprises taking a ranking of the set of source items into consideration in determining which item category or categories to filter out, said ranking being specific to the target user.

26. The non-transitory computer storage of claim 18, wherein the process is performed substantially in real time in response to a request by the target user for recommendations.