System and method for generating an alternative product recommendation

• US 20050004880A1
• Filed: 05/06/2004
• Published: 01/06/2005
• Est. Priority Date: 05/07/2003
• Status: Active Grant

• Current Assignee: CBS Interactive Inc. (Paramount Global (f/k/a ViacomCBS Inc.))
• Original Assignee: CNET Networks Incorporated (Paramount Global (f/k/a ViacomCBS Inc.))
• Inventors: Musgrove, Timothy A., Walsh, Robin Hiroko
• Granted Patent: US 7,783,528 B2
• Time in Patent Office: Days
• US Class Current: 705/400
• CPC Class Codes: G06Q 30/0253 During e-commerce, i.e. online transactions
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 30/0633 Lists, e.g. purchase orders, compilation or processing
• CPC Class Codes: G10L 21/00 Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility G10L19/00 takes precedence

Abstract

A method and system for automatically generating a naturally reading narrative product summary including assertions about a selected product. In one embodiment, the method includes the steps of determining at least one attribute associated with the selected product, selecting an alternative product, retrieving assertion models to describe the product and to recommend an alternative product in a natural manner, and generating a naturally reading narrative by combining the attribute and the selected alternative product with the retrieved assertion models such that the generated narrative includes a recommendation of the selected alternative product. A system is provided for implementing the described method.

Claims

1. A method of automatically generating a naturally reading narrative product summary including assertions about a selected product, said method comprising the steps of:

determining at least one attribute associated with said selected product, said at least one attribute including at least one of an attribute name, an attribute value, and an attribute unit;

selecting a comparable product;

retrieving assertion models defining forms in which assertions can be manifested to describe the selected product, and to identify said comparable product in a natural manner; and

generating a naturally reading narrative by combining said at least one attribute with a retrieved assertion model, and combining said selected comparable product with another retrieved assertion model to recommend said comparable product in said narrative.

2. The method of claim 1, further including the step of determining an attribute rank of said selected product and said selected comparable product, wherein said selected comparable product has an attribute rank that is substantially the same as said selected product for said at least one attribute.

3. The method of claim 1, further including the step of determining an attribute rank of said selected product and said selected comparable product, wherein said selected comparable product has an attribute rank that is higher than said selected product for said at least one attribute.

4. The method of claim 1, wherein said selected comparable product has a price that is substantially the same as a price of said selected product.

5. The method of claim 1, wherein said selected comparable product and said selected product have the same primary scenario.

6. The method of claim 1, wherein said selected comparable product has a price within a predetermined near-price margin of said selected product.

7. The method of claim 1, wherein said selected comparable product is ranked within a predetermined near-rank margin of said selected product.

8. A method of automatically generating a naturally reading narrative product summary including assertions about a selected product, said method comprising the steps of:

determining at least one attribute associated with said selected product, said at least one attribute including at least one of an attribute name, an attribute value, and an attribute unit;

selecting an alternative product;

retrieving assertion models defining forms in which an assertion can be manifested to describe the product, and to recommend said alternative product in a natural manner; and

generating a naturally reading narrative by combining said at least one attribute and said selected alternative product with said retrieved assertion models such that said generated narrative includes a recommendation of said selected alternative product.

9. The method of claim 8, wherein said selected alternative product has a price that is at least as high as price of said selected product.

10. The method of claim 8, wherein manufacturer of said selected alternative product is the same as said selected product.

11. The method of claim 8, wherein manufacturer of said selected alternative product is different than manufacturer of said selected product.

12. The method of claim 8, wherein said narrative includes at least one sentence recommending said selected alternative product.

13. The method of claim 8, wherein said recommendation is an advertisement, and further including the step of receiving compensation in exchange for recommending said selected alternative product.

14. The method of claim 13, wherein amount of said compensation is based on at least one of frequency of said recommendation and number of recommended products sold that is attributable to said recommendation.

15. The method of claim 8, wherein said assertion models comprise a plurality of assertion templates, each assertion template including a natural sentence pattern and at least one field corresponding to said at least one attribute.

16. The method of claim 15, wherein said at least one attribute is a plurality of attributes associated with said selected product and said selected alternative product, said method further including the step of assigning importance ratings to each of said plurality of attributes.

17. The method of claim 16, further including the step of designating at least one of said plurality of attributes as a key attribute, and wherein said generated naturally reading narrative is regarding said key attribute.

18. The method of claim 16, further including the step of determining attribute ranks for each of said plurality of attributes, and processing said attribute ranks together with said importance ratings to derive a severity value for each of said plurality of attributes.

19. The method of claim 18, wherein said step of deriving said severity values include the step of applying an inflection algorithm.

20. The method of claim 18, wherein said step of selecting said alternative product includes identification of a plurality of alternative products from a plurality of candidate alternative products.

21. The method of claim 20, further including the step of determining a severity differential between each of said plurality of attributes for said plurality of candidate alternative products and said selected product.

22. The method of claim 21, wherein said alternative product is selected based on at least one of a predetermined near-price margin, a predetermined near-rank margin, and a predetermined severity differential threshold.

23. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is within a predetermined near-price margin of said selected product;

has at least one attribute with a severity value that is above a predetermined severity differential threshold; and

has the same primary scenario as said selected product.

24. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold;

is priced lower and outside a predetermined near-price margin; and

has the same primary scenario as said selected product.

25. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is priced lower and outside a predetermined near-price margin; and

has the same primary scenario as said selected product.

26. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is within a predetermined near-price margin;

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold; and

has the same primary scenario as said selected product.

27. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is within a predetermined near-price margin;

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold; and

has a different primary scenario than said selected product.

28. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is priced higher and outside a predetermined near-price margin;

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is from the same manufacturer as said selected product; and

has the same primary scenario as said selected product.

29. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is priced higher and outside a predetermined near-price margin;

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is from the same manufacturer as said selected product; and

has the same primary scenario as said selected product.

30. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is priced higher than said selected product;

is within a predetermined near-rank margin for said key attribute; and

is from a different manufacturer as said selected product.

31. The method of claim 21, wherein said step of selecting said alternative product includes the step of identifying a candidate alternative product which:

is priced lower than said selected product;

is within a predetermined near-rank margin for said key attribute; and

is from the same manufacturer as said selected product.

32. A product summary generator for automatically generating a naturally reading narrative product summary including assertions about a selected product, said product summary generator comprising:

a product attribute module adapted to determine at least one attribute associated with said selected product, said attribute including at least one of an attribute name, an attribute value, and an attribute unit, and further adapted to select a comparable product;

an assertion model module adapted to retrieve assertion models that define forms in which assertions can be manifested to describe said selected product, and said selected comparable product in a natural manner; and

a summary generation module adapted to combine said attribute with a retrieved assertion model to describe said selected product in said narrative, and to combine said selected comparable product with another retrieved assertion model to recommend said selected comparable product in said narrative.

33. The product summary generator of claim 32, wherein said product attribute module determines an attribute rank of said selected product and said comparable product, said selected comparable product having an attribute rank that is substantially the same as said selected product for said at least one attribute.

34. The product summary generator of claim 32, wherein said product attribute module determines an attribute rank of said selected product and said selected comparable product, said selected comparable product having an attribute rank that is higher than said selected product for said at least one attribute.

35. The product summary generator of claim 32, wherein said selected comparable product has a price that is substantially the same as a price of said selected product.

36. The product summary generator of claim 32, wherein said selected comparable product and said selected product have the same primary scenario.

37. The product summary generator of claim 32, wherein said selected comparable product has a price within a predetermined near-price margin of said selected product.

38. The product summary generator of claim 32, wherein said selected comparable product is ranked within a predetermined near-rank margin of said selected product.

39. A product summary generator for automatically generating a naturally reading narrative product summary including assertions about a selected product, said product summary generator comprising:

a product attribute module adapted to determine at least one attribute associated with said selected product, said attribute including at least one of an attribute name, an attribute value, and an attribute unit;

an alternative product selection module adapted to select an alternative product;

an assertion model module adapted to retrieve assertion models that define forms in which an assertion can be manifested to describe said selected product, and recommend said selected alternative product, in a natural manner; and

a summary generation module adapted to generate a naturally reading narrative by combining said attribute and said selected alternative product with said retrieved assertion models so that said narrative includes a recommendation of said selected alternative product.

40. The product summary generator of claim 39, wherein said selected alternative product has a price that is at least as high as price of said selected product.

41. The product summary generator of claim 39, wherein manufacturer of said selected alternative product is the same as said selected product.

42. The product summary generator of claim 39, wherein manufacturer of said selected alternative product is different than manufacturer of said selected product.

43. The product summary generator of claim 39, wherein said narrative includes at least one sentence recommending said selected alternative product.

44. The product summary generator of claim 39, wherein said recommendation is an advertisement whereby compensation is provided for recommending said selected alternative product.

45. The product summary generator of claim 44, wherein amount of said compensation is based on at least one of frequency of said recommendation and number of recommended products sold that is attributable to said recommendation.

46. The product summary generator of claim 39, wherein said assertion models comprise a plurality of assertion templates, each assertion template including a natural sentence pattern and at least one field corresponding to said at least one attribute.

47. The product summary generator of claim 46, wherein said at least one attribute is a plurality of attributes associated with said selected product and said selected alternative product, said product attribute module assigns importance ratings to each of said plurality of attributes.

48. The product summary generator of claim 47, wherein said product attribute module designates at least one of said plurality of attributes as a key attribute, and wherein said generated naturally reading narrative is regarding said key attribute.

49. The product summary generator of claim 47, wherein said product attribute module determines attribute ranks for each of said plurality of attributes, and processes said attribute ranks together with said importance ratings to derive a severity value for each of said plurality of attributes.

50. The product summary generator of claim 49, wherein said product attribute module derives said severity values using an inflection algorithm.

51. The product summary generator of claim 49, further comprising an alternative products database for storing a plurality of candidate alternative products, wherein said alternative product selection module selects said alternative product from said alternative products database.

52. The product summary generator of claim 51, wherein said alternative product selection module determines a severity differential between each of said plurality of attributes for said plurality of candidate alternative products and said selected product.

53. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module based on at least one of a predetermined near-price margin, a predetermined near-rank margin, and a predetermined severity differential threshold.

54. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is within a predetermined near-price margin of said selected product;

has at least one attribute with a severity value that is above a predetermined severity differential threshold; and

has the same primary scenario as said selected product.

55. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold;

is priced lower and outside a predetermined near-price margin; and

has the same primary scenario as said selected product.

56. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is priced lower and outside a predetermined near-price margin; and

has the same primary scenario as said selected product.

57. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is within a predetermined near-price margin;

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold; and

has the same primary scenario as said selected product.

58. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is within a predetermined near-price margin;

is within a predetermined near-rank margin for said key attribute and for all attributes of said selected product that are above a predetermined high severity threshold; and

has a different primary scenario than said selected product.

59. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is priced higher and outside a predetermined near-price margin;

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is from the same manufacturer as said selected product; and

has the same primary scenario as said selected product.

60. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product in said alternative products database which:

is priced higher and outside a predetermined near-price margin;

has at least one attribute with a severity value that is above a predetermined severity differential threshold;

is from the same manufacturer as said selected product; and

has the same primary scenario as said selected product.

61. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product which:

is priced higher than said selected product;

is within a predetermined near-rank margin for said key attribute; and

is from a different manufacturer as said selected product.

62. The product summary generator of claim 52, wherein said alternative product is selected by said alternative product selection module by identifying a candidate alternative product which:

is priced lower than said selected product;

is within a predetermined near-rank margin for said key attribute; and

is from the same manufacturer as said selected product.

63. A method of automatically generating a naturally reading narrative product summary including assertions about a selected product, said method comprising the steps of:

determining at least one attribute associated with said selected product, said at least one attribute including at least one of an attribute name, an attribute value, and an attribute unit;

retrieving assertion models defining forms in which an assertion can be manifested to describe said selected product, and to recommend an alternative product in a natural manner;

receiving a plurality of bids from manufacturers that indicate an amount of compensation each manufacturer is willing to pay to have at least one of the manufacturer'"'"'s product be recommended as said alternative product;

selecting said alternative product based on said plurality of bids; and

generating a naturally reading narrative by combining said at least one attribute and said selected alternative product with said retrieved assertion models such that said generated narrative includes a recommendation of said selected alternative product.

64. The method of claim 63, wherein manufacturer of said selected alternative product is different than manufacturer of said selected product.

65. The method of claim 63, further including the step of receiving compensation in exchange for recommending said selected alternative product.

66. The method of claim 65, wherein amount of said compensation is based on at least one of frequency of said recommendation and number of recommended products sold that is attributable to said recommendation.

67. The method of claim 63, further including the step of establishing a limited sub-pool of products that may be selected for recommendation as an alternative product.