User interface and methods for recommending items to users

• US 7,720,723 B2
• Filed: 10/07/2002
• Issued: 05/18/2010
• Est. Priority Date: 09/18/1998
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Jacobi, Jennifer A., Benson, Eric A., Linden, Gregory D., Dicker, Russell A., Brownell, Jeffrey T.
• Primary Examiner(s): CHAMPAGNE, DONALD
• Application Number:
  US10/268,048
• Publication Number:
  US 20030105682A1
• Time in Patent Office: 2,780 Days
• Field of Search: 705/14, 705/27, 705/26
• US Class Current: 705/26.8
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0253 During e-commerce, i.e. online transactions
• CPC Class Codes: G06Q 30/0277 Online advertisement
• CPC Class Codes: G06Q 30/0633 Lists, e.g. purchase orders, compilation or processing

Abstract

An improved user interface and method are provided for presenting recommendations to a user when the user adds an item to a shopping cart. In response to the shopping cart add event, a page generation process generates and returns a page that includes a recommendations portion and a condensed view of the shopping cart. The recommendations portion preferably includes multiple recommendation sections, each of which displays a different respective set of recommended items selected according to a different respective recommendation or selection algorithm (e.g., recommendations based on shopping cart contents, recommendations based on purchase history, etc.). The condensed shopping cart view preferably lacks controls for editing the shopping cart, and lacks certain types of product information, making more screen real estate available for the display of the recommendations content. A link to a full shopping cart page allows the user to edit the shopping cart and view expanded product descriptions.

Claims

1. A computer-implemented method of recommending items to users of an electronic catalog system that provides electronic shopping carts for users, the method comprising:

receiving a request from a computing device of a user to add a selected item to a shopping cart;

in response to the request, selecting a plurality of item recommendation algorithms to use to provide item recommendations to the user, said plurality of item recommendation algorithms selected from a larger set of candidate item recommendation algorithms based at least partly on collected data associated with said user, at least some of the candidate item recommendation algorithms being operative to use different types of user activity than others to select items to recommend, wherein selecting said plurality of item recommendation algorithms comprises, for at least one candidate item recommendation algorithm, determining whether the collected data associated with said user is sufficient to use said candidate item recommendation algorithm;

executing each of the selected item recommendation algorithms such that each selected item recommendation algorithm selects a respective plurality of items to recommend to the user;

generating a first page that includes a plurality of recommendations sections and a shopping cart section, each recommendations section corresponding to, and identifying a plurality of items selected by, a respective one of the selected item recommendation algorithms, each recommendations section additionally including a respective label that corresponds to the respective item recommendation algorithm and including controls for adding the selected items to the shopping cart, said shopping cart section indicating the current contents of the shopping cart, said shopping cart section including a link to a second page that provides an option to edit the contents of the shopping cart via a set of editing screen elements, at least some of the editing screen elements not being included in the shopping cart section so that screen real estate is preserved on the first page for the display of the recommendations sections; and

returning the first page in response to the request to add the selected item to the shopping cart;

wherein the method is performed by a system that comprises one or more physical computers, and wherein generating said first page comprises, via execution of instructions by said system, generating said shopping cart section and populating each of said recommendations sections with identifiers of the selected items.

2. The method as in claim 1 further comprising generating and returning said second page in response to user selection of the link, wherein generating said second page comprises, for at least one item represented in the shopping cart, incorporating into the second page at least one type of item information that is not included on the first page, such that selection of the link provides user access to expanded item information for one or more items represented in the shopping cart.

3. The method as in claim 1 wherein at least one of the item recommendation algorithms selects items to recommend based at least-in part on a purchase history of the user.

4. The method as in claim 1 wherein at least one of the item recommendation algorithms selects items that, based on purchase histories of a population of users, are deemed to be related to the item selected to add to the shopping cart.

5. The method as in claim 1 wherein at least one of the item recommendation algorithms selects a set of items that, based on item viewing histories of a population of users, are deemed to be similar to the selected item.

6. The method as in claim 1, wherein the plurality of item recommendation algorithms includes a first recommendation algorithm that generates item recommendations based on the current contents of the shopping cart, and a second item recommendation algorithm that generates item recommendations without regard to the current contents of the shopping cart.

7. The method as in claim 1 wherein the method comprises selecting the plurality of item recommendation algorithms from said larger set of candidate item recommendation algorithms using an adaptive process that uses information regarding actions of past users, in combination with the collected data associated with the user, to predict which candidate item recommendation algorithms will be the most effective for said user.

8. The method of claim 1 wherein the method comprises selecting the plurality of item recommendation algorithms according to a hierarchy that gives some candidate item recommendations algorithms priority over others.

9. The method as in claim 1 wherein the method comprises constructing the first page such that the shopping cart section occupies less than one half of the width of the first page, and such that at least some of the recommendations sections are horizontally adjacent to the shopping cart section.

10. The method as in claim 9 wherein generating the first page comprises incorporating the shopping cart section into the page such that the shopping cart section occupies less than one third of the width of the first page.

11. The method as in claim 1 wherein generating the first page comprises, for each selected item recommendation algorithm, selecting at least two, and no more than five, items recommended by the respective item recommendation algorithm to represent on the first page, such that each recommendations section identifies two to five recommended items.

12. The method as in claim 1 wherein generating the first page comprises, for each selected item recommendation algorithm, selecting three items recommended by the respective item recommendation algorithm to represent on the first page, such that each recommendations section identifies exactly three recommended items.

13. The method of claim 1 wherein generating the first page comprises including, in at least a first recommendations section of said plurality of recommendations sections, a textual explanation of why the items represented in said first recommendations section were selected to recommend.

14. A method of recommending items to users of an electronic catalog system that provides electronic shopping carts for users, the method comprising:

receiving a request from a user computing device to add a selected item to a shopping cart of a user;

selecting a plurality of recommendation algorithms, from a set of candidate recommendation algorithms, to use to provide recommendations to the user in response to said request, such that the plurality of recommendation algorithms are selected based at least partly on an availability of user data used by particular candidate recommendation algorithms to generate recommendations; and

responding to the request by (a) executing the plurality of recommendation algorithms such that each recommendation algorithm selects a corresponding set of items to recommend to the user based on a different respective set of information regarding past activity of the user, and (b) generating and returning a first page that includes a shopping cart portion and a recommendations portion, said shopping cart portion specifying one or more items currently represented in the shopping cart (“

shopping cart contents”

), said recommendations portion including a plurality of recommendation sections, each of which corresponds uniquely to, and displays the set of items selected by, a respective one of the plurality of recommendation algorithms such that the user can distinguish between results of different recommendation algorithms, each recommendation section additionally including controls for adding individual recommended items to the shopping cart, said shopping cart and recommendations portions being visually demarcated on the first page, said shopping cart portion including a link to a second page that provides an option to edit the shopping cart contents via a set of editing screen elements, at least some of which are not included on the first page so that screen real estate is preserved on the first page for the display of the recommendations portion;

wherein the method, including receiving the request, selecting the plurality of recommendation algorithms, and responding to the request, is performed by a system that comprises one or more physical computers.

15. The method of claim 14 wherein the request is received while at least one additional item is already represented in the shopping cart, and the method comprises generating the first page such that the item selected to add is emphasized relative to the at least one additional item represented in the shopping cart.

16. The method as in claim 14 further comprising generating and returning said second page in response to selection of the link, wherein generating said second page comprises, for at least one item represented in the shopping cart, incorporating into the second page at least one type of item information that is not included on the first page, such that the link provides user access to expanded item information for the shopping cart contents.

17. The method as in claim 14 wherein the plurality of recommendation algorithms include a first recommendation algorithm that generates item recommendations based on the shopping cart contents, and a second recommendation algorithm that generates item recommendations without regard to the shopping cart contents.

18. The method as in claim 14 wherein at least one of the recommendation algorithms selects items that, based on purchase histories of a population of users, are deemed to be related to the item selected to add to the shopping cart.

19. The method as in claim 14 wherein at least one of the recommendation algorithms selects items that, based on item viewing histories of a population of users, are deemed to be similar to the item selected to add to the shopping cart.

20. The method as in claim 14 wherein the method comprises generating the first page such that the shopping cart portion occupies less than one half of a width of the first page, and such that the recommendations portion substantially occupies a remainder of said width of the first page.

21. The method as in claim 14 wherein the method comprises generating the first page such that the shopping cart portion occupies less than one third of the width of the first page.

22. The method of claim 14 wherein generating the first page comprises including, in at least a first recommendation section of said plurality of recommendation sections, a textual explanation of why the items represented in said first recommendation section are being recommended.

23. A computer-implemented method of providing item recommendations for a user, said method performed by a machine that comprises one or more physical computers, the method comprising:

executing a first recommendation algorithm to generate a first set of item recommendations for the user based on a first set of information reflective of item preferences of the user, said first set of item recommendations comprising identifiers of a first plurality of recommended items;

executing a second recommendation algorithm to generate a second set of item recommendations for the user based on a second set of information reflective of item preferences of the user, said second set of item recommendations comprising identifiers of a second plurality of recommended items; and

generating a page for presentation to the user, wherein generating the page comprises (1) populating a first recommendations section of the page with a representation of the first set of item recommendations such that each recommended item is presented with an associated control for adding the respective item to a shopping cart, (2) populating a second recommendations section of the page with a representation of the second set of item recommendations, and (3) generating a shopping cart section that identifies one or more items represented in the shopping cart and which includes a selectable control for proceeding to checkout, said first recommendations section, second recommendations section and shopping cart section being visually demarcated.

24. The method of claim 23 wherein the first recommendation algorithm generates item recommendations for the user based on current contents of the shopping cart of the user, and the second recommendation algorithm generates item recommendations for the user without regard to the current contents of the shopping cart.

25. The method of claim 23 further comprising dynamically selecting said first and second recommendation algorithms from a set of candidate recommendation algorithms based at least partly on what types of information are available for generating recommendations for the user.

26. The method of claim 23 wherein the method is performed in response to receiving a request to add an item to the shopping cart.

27. The method of claim 23 wherein generating said page comprises including, in each of the first and second recommendation sections, a respective textual explanation of why the items represented therein are being recommended.

28. A server system comprising one or more physical computers, said server system operative to perform a method that comprises:

executing a first recommendation algorithm to generate a first set of item recommendations for a user based on a first set of information reflective of item preferences of the user, said first set of item recommendations comprising identifiers of a first plurality of recommended items;

executing a second recommendation algorithm to generate a second set of item recommendations for the user based on a second set of information reflective of item preferences of the user, said second set of item recommendations comprising identifiers of a second plurality of recommended items; and

generating a page for presentation to the user, wherein generating the page comprises (1) populating a first recommendations section of the page with a representation of the first set of item recommendations such that each recommended item is presented with an associated control for adding the respective item to a shopping cart maintained by said server system, (2) populating a second recommendations section of the page with a representation of the second set of item recommendations such that each recommended item is presented with an associated control for adding the respective item to the shopping cart, and (3) generating a shopping cart section that identifies one or more items represented in the shopping cart and which includes a selectable control for proceeding to checkout, said first recommendations section, second recommendations section and shopping cart section being visually demarcated.

29. The server system of claim 28 wherein the server system is additionally operative to select said first and second recommendation algorithms from a set of candidate recommendation algorithms based at least partly on what types of user information are available for generating recommendations for the user.

30. The server system of claim 28 wherein the server system is additionally operative to select said first and second recommendation algorithms from a set of candidate recommendation algorithms using an adaptive process that uses information regarding actions of past users, in combination with collected data associated with the user, to predict which candidate item recommendation algorithms will be the most effective for said user.

31. The server system of claim 28 wherein the server system is operative to perform said method in response to the user selecting an item to add to the shopping cart.

32. The server system of claim 28 wherein the server system is additionally operative to include, in each of the first and second recommendation sections, a respective textual explanation of why the items represented therein are being recommended.