Data mining of user activity data to identify related items in an electronic catalog

• US 7,685,074 B2
• Filed: 11/19/2004
• Issued: 03/23/2010
• Est. Priority Date: 09/18/1998
• Status: Expired due to Term

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon.com, Inc.
• Inventors: Smith, Brent R., Linden, Gregory D., Zada, Nida K.
• Primary Examiner(s): Zeender; F. Ryan
• Assistant Examiner(s): ROJAS, HAJIME S
• Application Number:
  US10/994,089
• Publication Number:
  US 20050071251A1
• Time in Patent Office: 1,950 Days
• Field of Search: 705 26- 27, 705/1, 705/14
• US Class Current: 705/347
• 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 30/0254 based on statistics
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: G06Q 30/0271 Personalized advertisement
• CPC Class Codes: G06Q 30/0277 Online advertisement
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 30/0603 Catalogue ordering
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0633 Lists, e.g. purchase orders, compilation or processing
• CPC Class Codes: G06Q 30/0641 Shopping interfaces
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Various methods are disclosed for monitoring user browsing activities that indicate user interests in particular products, or other items, represented in an electronic catalog, and for using such information to identify items that are related to one another. In one embodiment, relationships between items within an electronic catalog are determined by identifying items that are frequently viewed by users within the same browsing session (e.g., items A and B are related because a significant portion of those who viewed A also viewed B). The resulting item relatedness data may be stored in a table that maps items to sets of related items. The table may be used to provide personalized item recommendations to users, and/or to supplement item detail pages of the electronic catalog with lists of related items. In one embodiment, the table is used to provide session-specific item recommendations to users.

Claims

1. A data mining method, comprising:

by a computer system that comprises one or more machines;

maintaining, in computer storage, session records of a plurality of users of an electronic catalog of items, said session records identifying catalog items selected by users for viewing within corresponding user sessions, said session records maintained by said computer system without requiring the users to explicitly create lists of items, wherein maintaining said session records comprises monitoring user accesses to item detail pages, each of which predominantly contains information about one particular catalog item, to thereby identify particular catalog items selected by users for viewing in said electronic catalog;

programmatically analyzing the session records to generate data values reflective of item co-occurrences within the session records, each data value corresponding to a respective pair of catalog items and representing a strength of a relationship between the two catalog items of said pair, each data value being dependent upon a number of said user sessions in which the item detail paces of both catalog items of the respective pair were accessed;

generating a data structure that associates particular catalog items with corresponding sets of related catalog items, wherein the data structure is generated based, at least in-part, on said data values; and

for each of a plurality of catalog items, using said data structure to supplement a corresponding item detail pace of the electronic catalog with a notification of other catalog items that are viewed by users who view the respective catalog item.

2. The method of claim 1, wherein the electronic catalog is arranged such that each user access to an item detail page generally represents an affirmative request by the user for information regarding a corresponding catalog item.

3. The method of claim 1, wherein analyzing the session records comprises determining, for an item pair consisting of a first catalog item and a second catalog item, a number of sessions in which respective item detail pages of the first and the second catalog items were both accessed.

4. The method of claim 1, wherein the catalog items are products represented in the electronic catalog.

5. The method of claim 1, wherein generating the data structure comprises analyzing the data values to determine which catalog items are to be mapped to each other in the data structure.

6. The method of claim 5, wherein generating the data structure further comprises storing the data values associated with the selected pairs of catalog items in the data structure.

7. The method of claim 1, further comprising, via execution of instructions by said computer system, using the data structure to generate personalized item recommendations for each of a plurality of target users.

8. The method of claim 1, further comprising, via execution of instructions by said computer system, using the data structure, in combination with a record of a plurality of catalog items viewed by a target user during a current session, to generate personalized, session-specific item recommendations for the target user.

9. The method of claim 1, wherein the method comprises supplementing an item detail page for a first catalog item with a list of additional catalog items, and with an indication that the additional catalog items in said list are viewed by users who view the first catalog item, said list being based on the data structure.

10. The method of claim 1, wherein each user session is a period of substantially continuous browsing activity by corresponding user, and each session record corresponds uniquely to a respective user session, said browsing activity being limited to actions performed during browsing of the electronic catalog.

11. The method of claim 1, wherein maintaining said session records comprises the computer system determining whether a particular session has ended based at least partly on whether a checkout transaction has occurred.

12. The method of claim 1, wherein the method comprises treating a user access to an item detail page of said electronic catalog as a selection of a corresponding catalog item for viewing.

13. The method of claim 1, wherein programmatically analyzing the session records comprises determining a distance between two catalog item viewing events within a session, and taking said distance into consideration in generating a data value for a corresponding pair of catalog items.

14. The method of claim 13, wherein determining said distance comprises determining a number of page accesses that occurred between the two catalog item viewing events.

15. The method of claim 13, wherein determining said distance comprises determining an amount of time between the two catalog item viewing events.

16. The method of claim 1, wherein the computer system is a web site system that comprises multiple machines.

17. A data mining system, comprising:

a computer system that comprises one or more machines, said computer system programmed, via executable code stored in computer storage, to perform a method that comprises;

maintaining, in computer storage, session records of a plurality of users of an electronic catalog of items, said session records identifying catalog items selected by users for viewing within corresponding user sessions, said session records maintained by said computer system without requiring the users to explicitly create lists of items, wherein maintaining said session records comprises monitoring user accesses to item detail pages, each of which predominantly contains information about one particular catalog item, to thereby identify particular catalog items selected by users for viewing in the electronic catalog;

analyzing the session records to generate data values reflective of item co-occurrences within the session records, each data value corresponding to a respective pair of catalog items and representing a strength of a relationship between the two catalog items of said pair, each data value being dependent upon a number of said user sessions in which the item detail pages of both catalog items of the respective pair were accessed;

generating a data structure that associates particular catalog items with corresponding sets of related catalog items, wherein the data structure is generated based, at least in-part, on said data values; and

for each of a plurality of catalog items, using said data structure to supplement a corresponding item detail page of the electronic catalog with a notification of other catalog items that are viewed by users who view the respective catalog item.

18. The data mining system of claim 17, wherein the computer system is programmed to analyze the session records in part by determining, for an item pair consisting of a first catalog item and a second catalog item, a number of sessions in which respective item detail pages of the first and second catalog items were both accessed.

19. The data mining system of claim 17, wherein the catalog items are products represented in the electronic catalog.

20. The data mining system of claim 17, wherein the computer system is programmed to use the data values to determine which catalog items are to be mapped to each other in the data structure.

21. The data mining system of claim 17, wherein the computer system is additionally programmed to use the data structure to generate personalized item recommendations for each of a plurality of target users.

22. The data mining system of claim 17, wherein the computer system is additionally programmed to use the data structure, in combination with a record of a plurality of catalog items viewed by a target user during a current session, to generate personalized, session-specific item recommendations for the target user.

23. The data mining system of claim 17, wherein the electronic catalog is arranged such that each user access to an item detail page generally represents an affirmative request by the user for information regarding a corresponding catalog item.

24. The data mining system of claim 17, wherein each session record represents a respective period of substantially continuous browsing activity by a user, said browsing activity being limited to actions performed during browsing of the electronic catalog.

25. The data mining system of claim 17, wherein the computer system is programmed to determine, in generating said session records, whether a particular session has ended based at least partly on whether a checkout transaction has occurred.

26. The data mining system of claim 17, wherein the computer system is programmed to (1) use the data structure to supplement an item detail page for a first catalog item with a list of additional catalog items, and (2) provide a notification that the catalog items in said list are viewed by users who view the first catalog item.

27. The data mining system of claim 17, wherein the computer system is programmed to measure a distance between two catalog item viewing events within a session, and to take said distance into consideration in generating a data value for a corresponding pair of catalog items.

28. The data mining system of claim 27, wherein the computer system is programmed to determine said distance based at least partly on a number of page accesses that occur between the two catalog item viewing events.

29. The data mining system of claim 27, wherein the computer system is programmed to determine said distance based at least partly on an amount of time than transpires between the two catalog item viewing events.

30. Physical computer storage which stores executable code that, when executed by a computer system, causes the computer system to perform a method that comprises:

maintaining, in computer storage, session records of a plurality of users of an electronic catalog of items, said session records identifying catalog items selected by users for viewing within corresponding user sessions, said session records maintained without requiring the users to explicitly create lists of items, wherein maintaining said session records comprises monitoring user accesses to item detail pages, each of which predominantly contains information about one particular catalog item, to thereby identify particular catalog items selected by users for viewing;

analyzing the session records to generate data values reflective of item co-occurrences within the session records, each data value corresponding to a respective pair of catalog items and representing a strength of a relationship between the two catalog items of said pair, each data value being dependent upon a number of said user sessions in which the respective item detail pages of both catalog items of the respective pair were accessed;

generating a data structure that associates particular catalog items with corresponding sets of related catalog items, wherein the data structure is generated based, at least in-part, on said data values; and

for each of a plurality of catalog items, using said data structure to supplement a corresponding item detail page of the electronic catalog with a notification of other catalog items that are viewed by users who view the respective catalog item.

31. The physical computer storage of claim 30, wherein the physical computer storage additionally stores executable code that is capable of causing the computer system to use the data structure, in combination with a record of a plurality of catalog items viewed by a target user during a current session, to generate personalized, session-specific item recommendations for the target user.

32. The physical computer storage of claim 30, wherein the electronic catalog is arranged such that each user access to an item detail page generally represents an affirmative request by the user for information regarding a corresponding catalog item.

33. The physical computer storage of claim 30, wherein the physical computer storage additionally stores executable code that is capable of causing the computer system to determine, in generating said session records, whether a particular session has ended based at least partly on whether a checkout transaction has occurred.

34. The physical computer storage of claim 30, wherein the method comprises supplementing an item detail page for a first catalog item with a list of additional catalog items, and with an indication that the catalog items in said list are viewed by users who view the first catalog item, said list being based on the data structure.

35. The physical computer storage of claim 30, wherein the session records are maintained in computer storage without any information that links the session records to corresponding users, whereby user privacy is maintained.

36. The method of claim 1, wherein maintaining the session records comprises storing the session records in said computer storage without any information linking the session records to corresponding users, whereby user privacy is maintained.

37. The data mining system of claim 17, wherein the computer system is programmed to store and analyze the session records anonymously.