Automatic item categorizer

• US 8,489,603 B1
• Filed: 10/23/2009
• Issued: 07/16/2013
• Est. Priority Date: 10/23/2009
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Europe Holdings Technologies SCS (Amazon.com, Inc.)
• Inventors: Weissgerber, Peter, Reuther, Patrick Oliver Henning, Kalampokis, Carlos David
• Primary Examiner(s): Nguyen, Cindy
• Application Number:
  US12/605,260
• Time in Patent Office: 1,362 Days
• Field of Search: 707/737, 707/767, 707/771, 707/804
• US Class Current: 707/737
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products

Abstract

A categorization service is described that automatically categorizes items of interest to a user. The user may possess an item that they wish to offer for sale using the network-based service. The user may prepare an electronic description of the item and submit this item information to a categorization service for categorization. Upon receipt, the categorization service may compare the item information to item category descriptions representing item categories of the network-based server in order to determine category recommendations for the item of interest. The recommended categories are returned to the user to enable the user to assign a category, selected from the category recommendations, to the item of interest.

Claims

1. A system for categorizing items, the system comprising:

a data store that stores item information for items;

a computing device operating a network-based service and in communication with the data store, wherein the network-based service is operative to;

receive a request for at least one category recommendation for an item of interest, the request including item information regarding the item of interest, wherein the item of interest is an item offered for sale using the network-based service; and

submit the received item information to a categorization service; and

the categorization service in communication with the network-based service, wherein the categorization service is operative to;

generate an item vector from the item information according to a vector space model;

compare the item vector to at least one category vector representing a description of an item category according to the vector space model, wherein the item category is maintained by the network-based service and associated with a plurality of items;

determine the item category to be the at least one category recommendation if a similarity between the item vector and the at least one category vector is greater than a threshold value; and

distribute the identified at least one category recommendation so as to enable a human reviewer to electronically select an item category from the at least one category recommendation for assignment to the item of interest.

2. The system of claim 1, wherein the categorization service is further operative to transmit updated item information for the item of interest to the network-based service, the updated item information comprising the item information and the assigned item category.

3. The system of claim 1, wherein the network-based service is further operative to store the updated item information in the data store.

4. The system of claim 1, wherein the item information comprises at least one of a flat file, an XML file, and item information directly input from a user.

5. The system of claim 1, wherein the similarity between the item vector and the at least one category vector comprises at least one of the deviation of the angles of the respective vectors and the cosine of the angles of the respective vectors.

6. A system for categorizing items of interest, the system comprising:

a data store that maintains a mathematical representation of a description of at least one item category; and

a computing device operating a categorization service and in communication with the data store, wherein the categorization service is operative to;

automatically determine at least one category recommendation for at least one item of interest offered for sale using a network-based service by;

generating a mathematical representation of item information associated with the at least one item of interest;

comparing the mathematical representations of the item information and the at least one item category description maintained in the data store, wherein the at least one item category description is maintained by the categorization service and is associated with a plurality of items; and

determining an item category to be the at least one category recommendation based upon a similarity of the mathematical representations of the item description and item category description; and

enable assignment of an item category to the item of interest, the item category selected from the at least one category recommendation.

7. The system of claim 6, wherein the item information is mathematically represented as an item vector represented according to a vector space model and the item category description is mathematically represented as a category vector according to the vector space model.

8. The system of claim 7, wherein the item category represented by the category vector is automatically determined to be the at least one category recommendation if the similarity between the item vector and the at least one category vector is greater than a threshold value.

9. The system of claim 8, wherein the similarity between the item vector and the at least one category vector comprises at least one of the deviation of the angles of the respective vectors and the cosine of the angles of the respective vectors.

10. The system of claim 6, wherein enabling assignment of an item category to the item of interest comprises enabling automatic selection of an item category from the at least one category recommendation.

11. The system of claim 6, wherein enabling assignment of an item category to the item of interest comprises enabling selection of the item category by a human reviewer.

12. The system of claim 11, wherein enabling selection of the item category by a human reviewer comprises providing the at least one category recommendation to at least one computing device operative to:

present the at least one category recommendation to a human reviewer; and

receive an item category assignment selected by the human reviewer from the at least one category recommendation.

13. The system of claim 6, wherein the categorization service is further operative to receive updated item information comprising the item information and the assigned item category.

14. The system of claim 13, wherein the categorization service is further operative to transmit the updated item information to a computing device in communication with another data store.

15. The system of claim 14, wherein the user comprises at least one of a manufacturer, wholesaler, or retailer of the item of interest.

16. The system of claim 6, wherein the categorization service is further operative to generate the category vector maintained by the data store.

17. The system of claim 6, wherein the received item information comprises at least one of an item title and an item description.

18. The system of claim 6, wherein the received item information comprises at least one of a flat file, an XML file, and item information directly input from a user.

19. The system of claim 6, wherein the at least one category recommendation comprises a null value.

20. A computer-implemented method for categorizing, on behalf of a user, items of interest in an item category maintained by a network-based service, the method comprising:

under control of one or more configured computer systems;

comparing a mathematical description of item information related to an item of interest to the user to a mathematical description of at least one description for an item category, wherein the item category is maintained by the network-based service and is associated with a plurality of items, and wherein the item of interest is an item offered for sale using a network-based service;

automatically determining at least one category recommendation for the item of interest based upon a similarity between the mathematical descriptions of the item information and at least one item category description; and

enabling selection of an item category from the at least one category recommendation for assignment to the item of interest.

21. The computer-implemented method of claim 20, wherein item information is mathematically represented as an item vector according to a vector space model and the item category description is mathematically represented as a category vector represented according to the vector space model.

22. The computer-implemented method of claim 21, wherein the item category represented by the item category vector is automatically determined to be the at least one category recommendation if the similarity between the item vector and the at least one category vector is greater than a threshold value.

23. The computer-implemented method of claim 22, wherein the similarity between the item vector and the at least one category vector comprises at least one of the deviation of the angles of the respective vectors and the cosine of the angles of the respective vectors.

24. The computer-implemented method of claim 20, wherein enabling selection of an item category comprises enabling automatic selection of an item category from the at least one category recommendation.

25. The computer-implemented method of claim 20, wherein enabling assignment of an item category comprises enabling selection of the item category by a human reviewer.

26. The computer-implemented method of claim 25, wherein enabling selection of the item category by a human reviewer comprises:

generating a user interface that enables a human reviewer to select an item category from the at least one category recommendation; and

transmitting the user interface and the at least one category recommendation to another computing device for display to the human reviewer.

27. The computer-implemented method of claim 20, wherein the method further comprises receiving updated item information comprising the received item information and the category assignment.

28. The computer-implemented method of claim 27, further comprising transmitting the updated item data to a computing device in communication with a data store.

29. The computer-implemented method of claim 20, wherein the received item information comprises at least one of an item title and an item description.

30. The computer-implemented method of claim 20, wherein the received item information comprises at least one of a flat file, an XML file, and item information directly input by a user.

31. The computer-implemented method of claim 20, wherein the at least one category recommendation comprises a null value.

32. A non-transitory computer-readable medium having encoded thereon instructions, wherein the instructions, when executed by a computing apparatus, cause the computing apparatus to:

generate a first user interface for receiving, from a user, item information regarding at least one item of interest to the user, wherein the item of interest is an item offered for sale using a network-based service;

automatically determine at least one category recommendation from vector representations of the received item information and at least one description for an item category maintained by a network-based service, wherein an item category is associated with a plurality of items and is determined to be a category recommendation if a similarity between the vector representation of the item information and the item category description is less than a threshold value; and

generate a second user interface that enables a human reviewer to assign an item category to the item of interest to the user, the item category selected from the at least one category recommendation.

33. The non-transitory computer-readable medium of claim 32, wherein the item information is mathematically represented as an item vector according to a vector space model and the item category description is mathematically represented as a category vector according to the vector space model.

34. The non-transitory computer-readable medium of claim 33, wherein the similarity between the item vector and the at least one category vector comprises at least one of the deviation of the angles of the respective vectors and the cosine of the angles of the respective vectors.

35. The non-transitory computer-readable medium of claim 32, wherein the second user interface further enables the received item information to be updated to include the item category assignment.

36. The non-transitory computer-readable medium of claim 35, wherein the categorization service is further operative to transmit the updated item information to a computing device in communication with another data store that maintains an account for a user associated with the item of interest.

37. The non-transitory computer-readable medium of claim 36, wherein the user comprises at least one of a manufacturer, wholesaler, or retailer of the item of interest.

38. The non-transitory computer-readable medium of claim 32, wherein the categorization service is further operative to generate the category vector.

39. The non-transitory computer-readable medium of claim 32, wherein the received item information comprises at least one of an item title and an item description.

40. The non-transitory computer-readable medium of claim 32, wherein the received item information comprises at least one of a flat file, an XML file, and item information directly input by a user.

41. The non-transitory computer-readable medium of claim 32, wherein the at least one category recommendation comprises a null value.