Dynamically generating digital product notifications over time based on product expiration data

Patent metadata

Field	Value
Publication number	US-11017422-B2
Application number	US-201815909714-A
Country	US
Kind code	B2
Filing date	Mar 1, 2018
Priority date	Mar 1, 2018
Publication date	May 25, 2021
Grant date	May 25, 2021

Abstract

Methods, systems, and non-transitory computer readable storage media are disclosed for dynamically generating discounted product digital notifications based on remaining product shelf life. For example, in one or more embodiments, the disclosed system determines an expiration date or a target product available for purchase from a merchant. Additionally, in one or more embodiments, the disclosed system utilizes a machine-learning model to dynamically generate discount prices for the target product over time based on the expiration date. In one or more embodiments, the disclosed system identifies a client device of a customer and provides a discount price for the target product for a given time window to the client device of the customer.

First claim

What is claimed is:

1. A non-transitory computer readable storage medium comprising instructions that, when executed by at least one processor, cause a computing device to:
determine an expiration date of a target product available for purchase from a merchant;
dynamically generate discount prices for the target product over time by processing both the expiration date of the target product and customer history data of a plurality of customers associated with the merchant utilizing learned parameters of a neural network to:
determine a plurality of probabilities of sale for the target product at a plurality of possible discount prices; and
select the discount prices for the target product from the plurality of possible discount prices based on the plurality of probabilities of sale, the discount prices comprising a first discount price for the target product generated based on the expiration date and customer history data for a first customer associated with the merchant;

identify a client device of the first customer associated with the merchant;
detect that a location of the client device of the first customer associated with the merchant enters a geo-fence corresponding to a location of the merchant; and
provide, to a client device of a first customer associated with the merchant and based on the customer history data for the first customer, the first discount price for the target product in response to the location of the client device entering the geo-fence.

2. The non-transitory computer readable storage medium as recited in claim 1 , further comprising instructions that, when executed by the at least one processor, cause the computing device to:
identify a client device of a second customer associated with the merchant, wherein the discount prices comprise a second discount price for the target product generated based on the expiration date and customer history data for the second customer; and
provide, to the client device of the second customer, the second discount price for the target product.

3. The non-transitory computer readable storage medium as recited in claim 1 , wherein the first discount price for the target product corresponds to a first time window, the non-transitory computer readable storage medium further comprising instructions that, when executed by the at least one processor, cause the computing device to provide a second discount price for the target product, the second discount price corresponding to a second time window.

4. The non-transitory computer readable storage medium as recited in claim 1 , further comprising instructions that, when executed by the at least one processor, cause the computing device to determine the expiration date of the target product by utilizing one or more sensors to determine a remaining shelf life of the target product, wherein the one or more sensors comprise an in-store camera or a camera on the client device of the first customer.

5. A computer-implemented method of dynamically generating discounted product digital notifications based on remaining product shelf life comprising:
determining, by at least one processor, expiration dates of a plurality of target products from a plurality of product categories available for purchase from a merchant;
dynamically generating, by the at least one processor, discount prices for the plurality of target products over time by processing both the expiration dates of the plurality of target products and customer history data of a plurality of customers associated with the merchant utilizing learned parameters of a neural network to:
determine a plurality of probabilities of sale for the plurality of target products at a plurality of possible discount prices; and
select the discount prices for the plurality of target products from the plurality of possible discount prices based on the plurality of probabilities of sale, the discount prices comprising a first discount price for a first target product generated based on a first expiration date and customer history data for a first customer;

identifying, by the at least one processor, a client device of the first customer associated with the merchant;
detecting, by the at least one processor, that a location of the client device of the first customer associated with the merchant enters a geo-fence corresponding to a location of the merchant; and
providing, to the client device of the first customer associated with the merchant and based on the customer history data for the first customer, the first discount price for the first target product in response to the location of the client device of the first customer entering the geo-fence.

6. The computer-implemented method as recited in claim 5 , wherein the discount prices comprise a second discount price for the first target product based on the first expiration date and customer history data for a second customer, the computer-implemented method further comprising:
identifying a client device of the second customer associated with the merchant; and
providing, to the client device of the second customer, the second discount price for the first target product.

7. The computer-implemented method as recited in claim 5 , further comprising determining the first expiration date of the first target product by utilizing one or more sensors to determine a remaining shelf life of the first target product, wherein the one or more sensors comprise an in-store camera or a camera on the client device of the first customer.

8. The computer-implemented method as recited in claim 5 , wherein the customer history data comprises user data profiles associated with the plurality of customers.

9. The computer-implemented method as recited in claim 5 , wherein dynamically generating the discount prices for the plurality of target products over time comprises generating, utilizing the learned parameters of the neural network, a predicted loss to the merchant for a possible discount price of the plurality of possible discount prices based on a total cost for product items of the first target product in inventory, a cost per product unit, an original price for the target product, the possible discount price as a percentage of the original price, and a probability of sale corresponding to the possible discount price.

10. The computer-implemented method as recited in claim 9 , wherein dynamically generating the discount prices for the plurality of target products over time comprises generating the first discount price by selecting a possible discount price from the plurality of possible discount prices that results in a lowest predicted loss to the merchant.

11. The computer-implemented method as recited in claim 5 , wherein dynamically generating the discount prices for the plurality of target products over time further comprises processing a purchase history and a price history of the plurality of products from the plurality of product categories.

12. The computer-implemented method as recited in claim 5 , wherein dynamically generating the discount prices for the plurality of target products over time comprises:
dynamically generating a first plurality of discount prices for a first plurality of target products of a first product category at a first store associated with the merchant; and
dynamically generating a second plurality of discount prices for a second plurality of target products of the first product category at a second store associated with the merchant, the second plurality of discount prices being different than the first plurality of discount prices.

13. The computer-implemented method as recited in claim 5 , wherein dynamically generating the discount prices

Assignees

• Adobe Inc

Inventors

• Smith Kevin
• Michaelson Cameron

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N3/08

  Learning methods · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06Q30/0224 Primary

  based on user history · CPC title