Detecting product lines within product search queries

• US 10,394,816 B2
• Filed: 12/27/2012
• Issued: 08/27/2019
• Est. Priority Date: 12/27/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google LLC (Alphabet Inc.)
• Inventors: Datta, Ritendra
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Khoshnoodi, Fariborz
• Application Number:
  US13/728,720
• Publication Number:
  US 20140188934A1
• Time in Patent Office: 2,434 Days
• Field of Search: 707771, 707769
• CPC Class Codes: G06F 16/24553 of query operations

Abstract

Systems and methods can determine product lines product searches. One or more computing devices can receive a product query of search terms. The product query may be classified to identify a product category. A brand may be identified for the product query. The brand may be selected from a list of known brands for the product category. One or more unknown product line terms may be identified within the product query. A metric may be computed to indicate how well the unknown product line terms correspond to an actual product line within the brand. The metric may be compared to a specified threshold. The unknown product line terms may be designated as a new product line of the brand if the metric favorably compares to the specified threshold. A product search may be performed on the product query. Product search results may be returned according to the product search.

Claims

1. A computer-implemented method to determine new product lines from unknown search terms in product searches, comprising:

receiving, using one or more computing devices, a first product query comprising first search terms;

classifying, using the one or more computing devices, the first product query to identify a product category for the first product query;

identifying, using the one or more computing devices, a brand corresponding to the first product query, the brand being selected from a list of known brands for the product category and being based on one or more of the first search terms;

retrieving, using the one or more computing devices, a known attribute list for the product category;

identifying, using the one or more computing devices, one or more unknown terms within the first product query that do not correspond to the known attribute list for the product category;

classifying, using the one or more computing devices, the one or more unknown terms not corresponding to the known attribute list as unknown product line terms based on comparing the one or more terms against known product line templates comprising known product lines corresponding to known brands;

computing, using the one or more computing devices, a metric indicating a correspondence of the unknown product line terms to an actual product line within the brand, wherein computing the metric comprises determining a total number of other product queries comprising the unknown product line terms;

determining, using the one or more computing devices, that the metric exceeds a predefined threshold;

designating, using the one or more computing devices, the unknown product line terms as a new product line for the brand in response to determining that the metric exceeds the specified threshold;

performing, using the one or more computing devices, a product search for the new product line using the first product query to obtain first search results;

refining, using the one or more computing devices, the first search results based on the new product line to obtain second search results; and

returning, using the one or more computing devices, search results according to the product search for the new product line.

2. The computer-implemented method of claim 1, wherein computing the metric further comprises determining one or more frequencies in which other product queries comprise both the unknown product line terms and terms associated with the brand.

3. The computer-implemented method of claim 1, wherein computing the metric further comprises determining one or more frequencies in which other product queries comprise both the unknown product line terms and terms associated with brands other than the brand.

4. The computer-implemented method of claim 1, wherein computing the metric further comprises evaluating each of the unknown product line terms for a similarity to known product lines associated with the brand.

5. The computer-implemented method of claim 1, wherein computing the metric further comprises evaluating the unknown product line terms for typical product line term patterns.

6. The computer-implemented method of claim 1, wherein computing the metric further comprises evaluating each of the unknown product line terms for a similarity to known product lines associated with the brand.

7. The computer-implemented method of claim 1, wherein the search results comprise one or more links to one or more products.

8. The computer-implemented method of claim 1, wherein computing the metric further comprises evaluating the unknown product line terms for typical product line term patterns.

9. A system to determine new product lines from unknown search terms in product searches, comprising:

one or more computing devices that are configured to;

receive a first product query comprising first search terms;

classify the first product query to identify a product category for the first product query;

identify a brand corresponding to the first product query, the brand being selected from a list of known brands for the product category and being based on one or more of the first search terms;

retrieve a known attribute list for the product category;

identify one or more unknown terms within the first product query that do not correspond to the known attribute list for the product category;

classify the one or more unknown terms not corresponding to the known attribute list as unknown product line terms based on comparing the one or more terms against known product line templates comprising known product lines corresponding to known brands;

compute a metric indicating a correspondence of the unknown product line terms to an actual product line within the brand, wherein computing the metric comprises determining a total number of other product queries comprising the unknown product line terms;

determine that the metric exceeds a predefined threshold;

designate the unknown product line terms as a new product line for the brand in response to determining that the metric exceeds the predefined threshold;

perform a product search for the new product line using the first product query to obtain first search results;

refine the first search results based on the new product line to obtain second search results; and

return second search results according to the product search for the new product line.

10. The system of claim 9, wherein computing the metric further comprises determining a total number of other product queries comprising the unknown product line terms.

11. The system of claim 9, wherein computing the metric further comprises determining one or more frequencies in which other product queries comprise both the unknown product line terms and terms associated with the brand.

12. The system of claim 9, wherein computing the metric further comprises evaluating each of the unknown product line terms for a similarity to known product lines associated with the brand.

13. The system of claim 9, wherein the search results comprise one or more links to one or more products.

14. The system of claim 9, wherein computing the metric further comprises determining one or more frequencies in which other product queries comprise both the unknown product line terms and terms associated with brands other than the brand.

15. The system of claim 9, wherein computing the metric further comprises evaluating each of the unknown product line terms for a similarity to known product lines associated with the brand.

16. The system of claim 9, wherein computing the metric further comprises evaluating the unknown product line terms for typical product line term patterns.

17. A computer program product to determine new product lines from unknown search terms in product searches, comprising:

a non-transitory computer-readable medium having computer-executable program instructions embodied thereon that, when executed by one or more computing devices, cause the one or more computing devices to;

receive a first product query comprising first search terms;

classify the first product query to identify a product category for the first product query;

identify a brand corresponding to the first product query, the brand being selected from a list of known brands for the product category and being based on one or more of the first search terms;

retrieve a known attribute list for the product category;

identify one or more unknown product line terms within the first product query that do not correspond to the known attribute list for the product category;

classify the one or more unknown terms not corresponding to the known attribute list as unknown product line terms based on comparing the one or more terms against known product line templates comprising known product lines corresponding to known brands;

compute a metric indicating a correspondence of the unknown product line terms to an actual product line within the brand, wherein computing the metric comprises determining a total number of other product queries comprising the unknown product line terms;

determine that the metric exceeds a predefined threshold;

designate the unknown product line terms as a new product line for the brand in response to determining that the metric exceeds the predefined threshold;

perform a product search for the new product line to obtain first search results;

refine the search results based on the new product line to obtain second search results; and

return search results according to the product search for the new product line.

18. The computer program product of claim 17, wherein computing the metric further comprises evaluating the unknown product line terms for typical product line term patterns.

19. The computer program product of claim 17, wherein computing the metric further comprises determining a total number of other product queries comprising the unknown product line terms.

20. The computer program product of claim 17, wherein computing the metric further comprises determining one or more frequencies in which other product queries comprise both the unknown product line terms and terms associated with the brand.