Identification of product categories

Patent metadata

Field	Value
Publication number	US-10032184-B2
Application number	US-201514750198-A
Country	US
Kind code	B2
Filing date	Jun 25, 2015
Priority date	Jun 29, 2011
Publication date	Jul 24, 2018
Grant date	Jul 24, 2018

Abstract

Disclosed are various embodiments for identifying clusters of user accounts who have similar purchase histories and generating product assignments for products identified in the purchase histories. The clusters of user accounts may depict an affinity towards a first type of product while depicting an aversion to a second type of product. The first type of product and the second type of product may have one or more opposing attributes. Product assignments for the products may be generated based on attributes of the products, product categories associated with the products, or the clusters.

First claim

Therefore, the following is claimed:

1. A non-transitory computer-readable medium embodying a program executable in at least one computing device that causes the at least one computing device to at least:
obtain imagery of a first product captured by at least one camera in communication with the at least one computing device;
perform an image recognition analysis algorithm on the imagery to extract at least one product attribute that appears on the first product;
identify a first attribute of the at least one product attribute based at least in part on at least one known image appearing on a packaging of the first product, the at least one known image corresponding to a logo or trademark;
identify a plurality of user accounts associated with an affinity towards a first product and an aversion towards a second product, the first product having at least one opposing attribute relative to the second product;
associate the plurality of user accounts with a cluster, the cluster defining the plurality of user accounts that exhibit a common purchasing behavior, wherein the common purchasing behavior indicates that less than a threshold percentage of users associated with the user accounts purchased the second product;
identify a product category associated with the first product; and
generate a product assignment for the first product based at least in part on the first attribute and the cluster.

2. The non-transitory computer-readable medium of claim 1 , wherein the program identifies the plurality of user accounts by:
determining that individual ones of the plurality of user accounts have purchased the first product; and
determining that the individual ones of the plurality of user accounts have not purchased the second product.

3. The non-transitory computer-readable medium of claim 1 , wherein the program identifies the plurality of user accounts by determining that at least a first threshold percentage of purchases in a purchase history associated with individual ones of the plurality of user accounts includes the first product.

4. The non-transitory computer-readable medium of claim 1 , wherein the program identifies the plurality of user accounts by associating the cluster with the product category in response to identifying the product category associated with the first product.

5. The non-transitory computer-readable medium of claim 4 , wherein the program further generates a user interface configured to receive a modification of the product assignment associated with the first product.

6. A system, comprising:
a processor-based computing device; and
an application executable in the processor-based computing device, the application causing the processor-based computing device to at least:
obtain imagery of a first product by at least one camera in communication with the processor-based computing device;
perform an image recognition analysis algorithm on the imagery to extract at least one product attribute that appears on the first product;
identify a first attribute of the at least one product attribute based at least in part on at least one known image appearing on a packaging of the first product, the at least one known image corresponding to a logo or trademark;
identify a plurality of user accounts associated with the first product in response to determining that the subset of user accounts exhibit an affinity toward the first product and an aversion toward a second product, wherein the identified first attribute is an opposing attribute relative to the second product;
associate the subset of user accounts with a cluster defining the plurality of user accounts that exhibit a common purchasing behavior, wherein the common purchasing behavior indicates that less than a threshold percentage of users associated with the user accounts purchased the second product; and
generate a product assignment for the first product based at least in part on the first attribute and the cluster.

7. The system of claim 6 , wherein the application identifies the subset of user accounts by causing the processor-based computing device to at least:
determine whether at least one search term is associated with the first product; and
associate the cluster with the at least one search term.

8. The system of claim 6 , wherein the application generates the product assignment for the first product by causing the processor-based computing device to at least generate an alert transmitted to an administrative client device, the alert identifying the cluster and the first attribute in response to receiving a confirmation from the administrative client device to generate the product assignment in an inventory item that associates a product category with the first product.

9. The system of claim 6 , wherein the application further causes the processor-based computing device to at least:
identify the first attribute associated with the first product in response to extracting an indication of the first attribute from at least one image of a packaging of the first product;
identify at least one keyword associated with at least one inventory item corresponding to the first product; and
generate the product assignment for the first product in response to associating a product category with the first product based at least in part on the at least one keyword.

10. The system of claim 6 , wherein the application further causes the processor-based computing device to at least generate the product assignment in response to associating a product category with the first product based at least in part on at least one entry of at least one third party data feed associated with the first product and the cluster.

11. A method, comprising:
obtaining imagery of a first product captured by at least one camera in communication with one or more computing devices;
performing, via at least one of the one or more computing devices, an image recognition analysis algorithm on the imagery to extract at least one product attribute that appears on the first product;
identifying, via at least one of the one or more computing devices, a first attribute of the at least one product attribute based at least in part on at least one known image appearing on the product packaging, the at least one known image corresponding to a logo or trademark;
identifying, via at least one of the one or more computing devices, a plurality of user accounts associated with the first product, the plurality of user accounts exhibiting an affinity towards the first product and an aversion towards a second product, and the first attribute is an opposing attribute relative to the second product;
associating, via at least one of the one or more computing devices, the plurality of user accounts with a cluster defining the plurality of user accounts that exhibit a common purchasing behavior, wherein the common purchasing behavior indicates that less than a threshold percentage of users associated with the user accounts purchased the second product; and
generating, via at least one of the one or more computing devices, a product assignment for the first product based at least in part on the first attribute and the cluster.

12. The method of claim 11 , wherein identifying the plurality of user accounts associated with the first product further comprises determining that at least a first second threshold percentage of total purchases from a purchase history of individual ones of the plurality of user accounts is made from a product category associated with the first product.

13. The method of claim 11 , wherein the first attribute is further identified based at least in part on at least on text item extracted from an op

Assignees

• Amazon Tech Inc

Inventors

• Steves Andrea C
• Lehmann Sebastian

Classifications

• G06Q30/0255

  based on user history · CPC title

• G06Q30/0224 Primary

  based on user history · CPC title

• G06Q30/0201

  Market modelling; Market analysis; Collecting market data · CPC title