Aggregating product information for electronic product catalogs

• US 8,548,878 B1
• Filed: 03/11/2011
• Issued: 10/01/2013
• Est. Priority Date: 03/11/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Li, Nan, Graham, Michael
• Primary Examiner(s): Rosen, Nicholas D
• Application Number:
  US13/045,973
• Time in Patent Office: 935 Days
• Field of Search: 705/26.64, 705/27.1, 705/27.2
• US Class Current: 705/27.2
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06Q 30/0603 Catalogue ordering
• CPC Class Codes: G06V 10/762 using clustering, e.g. of similar faces in social networks

Abstract

A product catalog includes information regarding products for sale online by various merchants. An analysis software module can collect information regarding new product offers and associate that information with existing product information in the catalog. The module can identify one or more potential product matches in the catalog based on identical or similar strong identifier information between the offer and the potential match, and/or based on query search result relationships between the offer and the potential match. The analysis module can evaluate each potential match by performing an image comparison between one or more product images provided with the offer and one or more representative images selected by the analysis module for the potential matching product. If the analysis module confirms a match, the analysis module can add information regarding the new product offer to the product catalog, associating the new product offer with the confirmed matching product.

Claims

1. A computer-implemented method for aggregating product information for an electronic product catalog, comprising the steps of:

selecting, for each product in an electronic product catalog comprising information regarding products available for sale online, at least one representative image;

receiving information regarding a product offer, the received information comprising information identifying a product subject to the product offer and at least one image of the product subject to the product offer;

identifying, based on the received information, at least one product in the electronic product catalog that potentially matches the product subject to the product offer;

determining, by an analysis module, whether each identified product in the electronic product catalog actually corresponds to the product subject to the product offer by completing an image analysis on the received image and the at least one representative image for each identified product in the electronic product catalog; and

in response to determining that a particular identified product actually corresponds to the product subject to the product offer, adding, by the analysis module, information regarding the product offer to the electronic product catalog, the added information associated with the particular identified product in the electronic product catalog,

wherein the analysis module is implemented in at least one computer system that comprises instructions stored in a machine-readable medium and a processor that executes the instructions.

2. The method of claim 1, wherein the step of identifying the at least one product in the electronic product catalog that potentially matches the product subject to the product offer comprises the step of determining that the product offer comprises a strong identifier associated with each of the at least one product in the electronic product catalog.

3. The method of claim 1, wherein the step of identifying the at least one product in the electronic product catalog that potentially matches the product subject to the product offer comprises the step of determining, for respective ones of the at least one product in the electronic product catalog, that information regarding the product offer and

information regarding the respective product in the electronic product catalog were both included in a set of search results displayed in response to the same product search query.

4. The method of claim 1, wherein the step of selecting the at least one representative image comprises the steps of:

for each product in the electronic product catalog,

identifying images that are known to correspond to the product in the electronic product catalog;

measuring similarity among the identified images;

identifying clusters of similar identified images based on the measured similarity; and

selecting, from each identified cluster, the at least one representative image.

5. The method of claim 1, further comprising the step of not adding information regarding the product offer to the electronic product catalog in response to determining that none of the identified products in the electronic product catalog actually correspond to the product subject to the product offer.

6. The method of claim 5, further comprising the step of overriding the step of not adding information regarding the product offer to the electronic product catalog in response to determining that the identified product in the electronic product catalog does not actually correspond to the product subject to the product offer.

7. A computer-implemented method for aggregating product information for an electronic product catalog, comprising the steps of:

receiving information regarding a product offer, the received information comprising information identifying a product subject to the product offer and at least one image of the product subject to the product offer;

identifying, based on the received information, a product in an electronic product catalog that potentially matches the product subject to the product offer, the product in the electronic product catalog having at least one representative image associated therewith;

determining, by an analysis module, whether the identified product in the electronic product catalog actually corresponds to the product subject to the product offer by determining whether the received image matches the at least one representative image for the product in the electronic product catalog; and

adding, by the analysis module, information regarding the product offer to the electronic product catalog, the added information associated with the identified product in the electronic product catalog, in response to determining that the identified product in the electronic product catalog actually corresponds to the product subject to the product offer, wherein the analysis module is implemented in at least one computer system that comprises instructions stored in a machine-readable medium and a processor that executes the instructions.

8. The method of claim 7, wherein the step of identifying the product in the electronic product catalog that potentially matches the product subject to the product offer comprises the step of determining that the product offer comprises a strong identifier associated with the product in the electronic product catalog.

9. The method of claim 7, wherein the step of identifying the product in the electronic product catalog that potentially matches the product subject to the product offer comprises the step of determining that information regarding the product offer and information regarding the product in the electronic product catalog were both included in a set of search results displayed in response to the same product search query.

10. The method of claim 7, further comprising the step of selecting from a set of product images the at least one representative image for the identified product in the electronic product catalog.

11. The method of claim 10, wherein the step of selecting the at least one representative image comprises the step of performing an image analysis on at least a portion of the product images in the set of product images to identify similar images.

12. The method of claim 10, wherein the step of selecting the at least one representative image further comprises the steps of:

identifying images that are known to correspond to the identified product in the electronic product catalog;

using image analysis to measure similarity among the identified images;

identifying clusters of similar images based on the image analysis; and

selecting, from an identified cluster, the at least one representative image.

13. The method of claim 7, further comprising the step of not adding information regarding the product offer to the electronic product catalog in response to determining that the identified product in the electronic product catalog does not actually correspond to the product subject to the product offer.

14. The method of claim 13, further comprising the step of overriding the step of not adding information regarding the product offer to the electronic product catalog.

15. A computer program product, comprising:

a non-transitory computer-readable medium having computer-readable program code embodied therein for aggregating product information for an electronic product catalog, the computer-readable program code, which, when implemented by one or more processors, causes the one or more processors to carry out the steps of;

using near-duplicate image analysis to select, for each product in an electronic product catalog, at least one representative image;

receiving information regarding a product offer, the received information comprising information identifying a product subject to the product offer and at least one image of the product subject to the product offer;

identifying, based on the received information, at least one product in the electronic product catalog that potentially matches the product subject to the product offer;

determining whether each identified product in the electronic product catalog actually corresponds to the product subject to the product offer by completing an image analysis on the received image and the at least one representative image for each identified product in the electronic product catalog; and

in response to determining that a particular identified product actually corresponds to the product subject to the product offer, adding information regarding the product offer to the electronic product catalog, the added information associated with the particular identified product in the electronic product catalog.