System and methods for collaborative recommendations

• US 6,064,980 A
• Filed: 03/17/1998
• Issued: 05/16/2000
• Est. Priority Date: 03/17/1998
• Status: Expired due to Term

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon.com, Inc.
• Inventors: Jacobi, Jennifer A., Benson, Eric A.
• Primary Examiner(s): Voeltz, Emanuel Todd
• Assistant Examiner(s): McCarty, William R.
• Application Number:
  US09/040,171
• Time in Patent Office: 791 Days
• Field of Search: 705/26, 705/27, 705/1
• US Class Current: 705/26.7
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0641 Shopping interfaces

Abstract

A recommendation service is disclosed which uses collaborative filtering techniques to recommend books to users of a Web site. The Web site includes a catalog of the various titles that can be purchased via the site. The recommendation service includes a database of titles that have previously been rated and that can therefore be recommended by the service using collaborative filtering methods. At least initially, the titles and title categories (genres) that are included within this database (and thus included within the service) are respective subsets of the titles and categories included within the catalog. As users browse the site to read about the various titles contained within the catalog, the users are presented with the option of rating specific titles, including titles that are not currently included within the service. The ratings information obtained from this process is used to automatically add new titles and categories to the service. The breadth of categories and titles covered by the service thus grows automatically over time, without the need for system administrators to manually collect and input ratings data. To establish profiles for new users of the service, the service presents new users with a startup list of titles, and asks the new users to rate a certain number of titles on the list. To increase the likelihood that new users will be familiar with these titles, the service automatically generates the startup list by identifying the titles that are currently the most popular, such as the titles that have been rated the most over the preceding week.

Claims

1. A method of adding items and predefined item categories to a recommendation service that uses collaborative filtering to recommend items to users from a set of service items within a set of service categories, comprising:

providing users an option to rate items, including non-service-items and items in non-service-categories;

when a user rates a non-service-item that falls within a service category, adding the non-service-item to the service;

when a user rates a non-service-item that does not fall within any service category, recording the item and associated user rating in a repository without adding the item to the service; and

when a quantity of rated items recorded within the repository for a non-service-category reaches a level that represents a desired degree of reliability for recommending items within that category, adding the category and associated rated items to the service.

2. The method as in claim 1, wherein the items comprise book titles, and the item categories comprise book genres.

3. The method as in claim 1, wherein providing users an option to rate items comprises presenting users of the recommendation service modified web pages that include forms for rating items during browsing of an electronic catalog.

4. The method as in claim 1, wherein adding the category and associated rated items to the service comprises moving user ratings data for a plurality of items from the repository to a database of the recommendation service.

5. The method as in claim 1, wherein adding the category and associated rated items to the service is performed automatically in response to an items-per-category threshold being reached within the repository.

6. The method as in claim 1, wherein adding the category and associated rated items to the service is performed under the control of a system administrator.

7. The method as in claim 1, further comprising providing users an option to select a service category in which to obtain category-specific recommendations.

8. A method of adding predefined item categories to a recommendation service that uses collaborative filtering to recommend items to users from a set of service categories, comprising:

providing users an option to rate individual items, including items that do not fall within any service category;

when a user rates an item that does not fall within any service category, recording the item and associated user rating in a repository without adding the item to the service; and

when a quantity of rated items recorded within the repository for a non-service-category reaches a level that represents a desired degree of reliability for recommending items within that category, adding the non-service-category to the service.

9. The method as in claim 8, wherein the items comprise book titles, and the item categories comprise book genres.

10. The method as in claim 8, wherein providing users an option to rate items comprises presenting users of the recommendation service modified web pages that include forms for rating items during browsing of an electronic catalog.

11. The method as in claim 8, wherein adding the non-service-category to the service comprises moving user ratings data for a plurality of items from the repository to a database of the recommendation service.

12. The method as in claim 8, wherein adding the non-service-category to the service is performed automatically in response to an items-per-category threshold being reached within the repository.

13. The method as in claim 8, wherein adding the non-service-category to the service is performed under the control of a system administrator.

14. The method as in claim 8, further comprising providing users an option to select a service category in which to obtain category-specific recommendations.

15. In a computer-implemented collaborative recommendations service in which items are recommended to users from a database within a set of service categories, the set of service categories being a subset of the item categories of the database, a method of adding categories to the service, comprising:

(a) prompting users to rate items that fall outside the set of service categories;

(b) storing ratings data resulting from (a) without adding the rated items to the service; and

(c) when a quantity of ratings data stored in (b) for a non-service-category of the database reaches a selected level that represents a desired degree of reliability for recommending items in the non-service-category, adding the non-service-category to the service.

16. The method as in claim 15, wherein the items comprise book titles, and the item categories comprise book genres.

17. The method as in claim 15, wherein prompting users to rate items comprises presenting users of the recommendation service modified web pages that include forms for rating items during browsing of an electronic catalog.

18. The method as in claim 15, wherein adding the non-service-category to the service comprises moving user ratings data for a plurality of items from a repository to a database of recommendable items.

19. The method as in claim 15, wherein adding the non-service-category to the service is performed automatically in response to an items-per-category threshold being reached.

20. The method as in claim 15, wherein adding the non-service-category to the service is performed under the control of a system administrator.

21. The method as in claim 15, further comprising providing users an option to select a service category in which to obtain category-specific recommendations.

22. A system that implements a recommendation service, comprising:

a database of items that fall within a plurality of categories;

a recommendation service component which recommends items to users from the database within a subset of the plurality of categories ("service categories") using a collaborative filtering method; and

a service populating component which;

provides users an option to rate items, including non-service-items and items in non-service-categories;

when a user rates a non-service-item that falls within a service category, adds the non-service-item to the service; and

when a user rates a non-service-item that does not fall within any service category, records the item and associated user rating without adding the item to the service, to thereby collect data for adding non-service-categories to the service as desired rated-items per category levels are reached.

23. The system as in claim 22, wherein the service populating component automatically adds a non-service-category to the service when a rated-items per category threshold for that category is reached.

24. The system as in claim 22, wherein the items comprise book titles, and the predefined item categories comprise book genres.

25. A method for generating a ratings profile for use within a recommendation service that recommends items to users within multiple service categories, comprising:

prompting the user to specify a preference level for each of a plurality of the service categories;

using the preference levels specified by the user to generate a list of items that reflects the user'"'"'s category preferences; and

prompting the user to rate items from the list.

26. The method as in claim 25, wherein using the preference levels to generate a list comprises using the preference levels to filter a startup list.