System for recommending item bundles

• US 8,285,602 B1
• Filed: 11/19/2009
• Issued: 10/09/2012
• Est. Priority Date: 11/19/2009
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Yi, Jin, Levitan, Avi A.
• Primary Examiner(s): Smith, Jeffrey A
• Assistant Examiner(s): Desai, Resha
• Application Number:
  US12/622,270
• Time in Patent Office: 1,055 Days
• Field of Search: 705 261- 272
• US Class Current: 705/26.7
• CPC Class Codes: G06Q 30/00 Commerce

Abstract

This disclosure describes systems and methods for recommending bundles of items to users. A bundle can include an item that is substitutable for an item selected by a user and one or more related items that complement the substitutable item. For example, if a user has selected a wireless device on a page of a network application, a bundle recommendation can be provided that includes a similar wireless device and related accessories, a compatible service plan, and the like. Recommending bundles of items can be beneficial for providing users with access to items that they might like to purchase together, without requiring users to separately find each item.

Claims

1. A method of recommending a wireless device bundle having a wireless device and associated items to a user of an electronic catalog, the method comprising:

by a computing system that comprises one or more computing devices;

identifying a first wireless device from a plurality of items represented in an electronic catalog, the first wireless device associated with a first carrier;

identifying a substitute wireless device that has a substitutability association with the first wireless device, the substitute wireless device associated with a second carrier different from the first carrier, the substitutability association being based at least partly on a programmatic analysis of first behavioral data of one or more users of the electronic catalog with respect to the first wireless device and the substitute wireless device, the behavioral data comprising data regarding those of the items in the electronic catalog that were selected by a plurality of users;

identifying at least one complementary item having a complementary association with the substitute wireless device, the complementary association being based at least partly on a programmatic analysis of second behavioral data with respect to the substitute wireless device, the at least one complementary item comprising an accessory for the substitute wireless device; and

in response to a target user selecting a representation of the first wireless device for viewing, outputting a representation of a bundle of items configured to be selectable by the target user, the bundle of items identifying the substitute wireless device and the at least one complementary item, wherein said outputting comprises supplementing a product detail page for the first wireless device with a recommendation to purchase the bundle of items in combination, the bundle being selectable by the target user to cause the bundle to be added to an electronic shopping cart.

2. The method of claim 1, further comprising, in response to receiving a selection of the bundle of items by the target user, adding a representation of the substitute wireless device and the at least one complementary item to an electronic shopping cart of the target user.

3. The method of claim 1, wherein said programmatic analysis of second behavioral data comprises an analysis of one or more of item viewing and purchase actions of a plurality of users of the electronic catalog.

4. A computer-implemented method of recommending bundles of items, the method comprising:

by a computing system that comprises one or more computing devices;

identifying a substitutable item that has a substitution association with a first item selected from a plurality of items represented in an electronic catalog, the substitution association reflecting a first automated assessment of a degree to which the substitutable item is substitutable for the first item;

identifying one or more complementary items that each have a complementary association with the substitutable item, each of the complementary associations reflecting a second automated assessment of a degree to which the substitutable item is complementary to one of the one or more complementary items;

selecting a subset of the complementary items; and

combining the subset of the complementary items with the substitutable item to form a bundle of items; and

in response to a target user selecting one or more items from the electronic catalog, identifying the bundle of items as being related to the one or more selected items and outputting the bundle of items for recommendation to the user, wherein said outputting comprises supplementing an item detail page with a recommendation to purchase the bundle of items in combination, the bundle being selectable by the target user to cause the bundle to be added to an electronic shopping cart;

wherein the one or more selected items comprise a first wireless device associated with a first carrier and wherein the bundle of items comprises a second wireless device associated with a second carrier different from the first carrier.

5. The method of claim 4, wherein the electronic catalog comprises a hierarchical arrangement of categories, such that each of the plurality of items is assigned to at least one category.

6. The method of claim 5, wherein said identifying one or more complementary items comprises identifying each of the one or more complementary items from a different category.

7. The method of claim 4, further comprising determining whether the bundle of items is valid to be purchased by a user.

8. The method of claim 4, wherein said identifying one or more complementary items comprises identifying two or more items that each have a complementary association with each other.

9. The method of claim 4, wherein said identifying one or more complementary items comprises identifying two or more items that are not substitutable for each other.

10. The method of claim 4, wherein said identifying one or more complementary items comprises identifying one or more accessories for the substitutable item.

11. The method of claim 4, wherein said identifying one or more complementary items comprises identifying one or more services related to the substitutable item.

12. A system for mining bundles of associated items in order to recommend the bundles to users, the system comprising:

a substitution associations dataset comprising data representing substitution associations between a plurality of items represented in an electronic catalog, each of the substitution associations reflecting a degree to which selected ones of the items are substitutable for each other;

a complementary associations dataset comprising data representing complementary associations between items, each of the complementary associations reflecting a degree to which selected ones of the items complement each other; and

a bundle recommender comprising computer hardware, the bundle recommender operative to;

identify a substitutable item from the substitution association dataset that has a substitution association with a first item,

identify one or more complementary items from the complementary associations dataset that each have a complementary association with the substitutable item,

combine a subset of the complementary items with the substitutable item to form a bundle of items, and

in response to a target user selecting one or more items from the electronic catalog;

identify the bundle of items as being related to the one or more selected items and output the bundle of items for recommendation to the user, wherein the bundle recommender is configured to output the bundle of items by supplementing an item detail page with a recommendation to purchase the bundle of items in combination, the bundle being selectable by the target user to cause the bundle to be added to an electronic shopping cart;

wherein the one or more selected items comprise a first wireless device associated with a first carrier and wherein the bundle of items comprises a second wireless device associated with a second carrier different from the first carrier.

13. The system of claim 12, wherein the bundle recommender is further operative to output the representation of the bundle of items in response to the target user selecting a representation of the first item, the first item being one of the one or more selected items.

14. The system of claim 12, further comprising a bundle mining system configured to generate the substitution associations dataset at least in part by comparing attributes of the plurality of items represented in the electronic catalog.

15. The system of claim 12, further comprising a bundle mining system configured to generate the complementary associations dataset at least in part by at least in part by analyzing observed user behavior with respect to the plurality of items represented in the electronic catalog.

16. The system of claim 15, wherein the observed user behavior comprises item selection events.

17. The system of claim 16, wherein the item selection events comprise one or more of item viewing events and item purchase events.

18. The system of claim 12, wherein the first item is a wireless device, and wherein the one or more complementary items comprise at least one of a wireless plan, a wireless service option, and an accessory for the wireless device.