Methods and apparatus for automatically providing digital advertisements

Patent metadata

Field	Value
Publication number	US-11763349-B2
Application number	US-202016748753-A
Country	US
Kind code	B2
Filing date	Jan 21, 2020
Priority date	Jun 27, 2019
Publication date	Sep 19, 2023
Grant date	Sep 19, 2023

Abstract

This application relates to apparatus and methods for automatically determining and providing digital advertisements to targeted users. In some examples, a computing device receives campaign data identifying items to advertise on a website, and generates campaign user data identifying a user that has engaged all of the items on the website. The computing device may then determine a portion of the users based on a relationship between each user and the campaign user data, and may determine user-item values for each of the items for each user of the portion of users, where each user-item value identifies a relational value between the corresponding user and item. The computing device may then identify one or more of the items to advertise to each user of the portion of users based on the user-item values, and may transmit to a web server an indication of the items to advertise for each user.

First claim

What is claimed is:

1. A system comprising:
a server;
a database; and
a computing device comprising at least one processor and communicatively coupled to the database, the computing device configured to:
receive campaign data identifying a plurality of advertising campaigns for a website, each of the plurality of advertising campaigns identifying a plurality of campaign items;
generate, for each of the plurality of advertising campaigns, campaign user data based on the plurality of campaign items corresponding to each of the plurality of advertising campaigns, wherein the campaign user data for each of the plurality of advertising campaigns identifies a campaign user that has engaged with the plurality of campaign items for the corresponding advertising campaign;
obtain, from the database, user attribute data identifying at least one attribute of a user of the website;
obtain, from the database, at least one of user session data and user transaction data for the user;
determine a plurality of items based on the at least one of the user session data and the user transaction data;
obtain, from the database, item attributes for each of the plurality of items;
generate a product embedding average vector for the plurality of items, wherein each element of the product embedding average vector identifies an average value for an item attribute of the plurality of items;
generate a user vector for the user based on the product embedding average vector and the user attribute data;
apply a neural network to the user vector to generate user embedding data for the user, wherein the neural network is a stacked denoising auto-encoder trained based on:
inputting noisy user vectors of a plurality of users of the website to the neural network, and
minimizing a least squared loss between an output of the neural network and the noisy user vectors;

determine, for each of the plurality of advertising campaigns, a user campaign value based on the campaign user data and the user embedding data;
determine user engagement data characterizing user engagements of each of the plurality of items based on the at least one of user session data and the user transaction data;
determine a number of the plurality of advertising campaigns to be displayed to the user, wherein each respective advertising campaign of the number of the plurality of advertising campaigns is determined to be displayed to the user based on:
generating a user-item affinity value associated with each campaign item of the respective advertising campaign based on the user engagement data,
determining a total quantity of campaign items of the respective advertising campaign, wherein each of the total quantity of campaign items is associated with a user-item affinity value higher than a first threshold, and
determining that the total quantity is higher than a second threshold;

generate assignment data assigning to the user the number of the plurality of advertising campaigns based on the user campaign values and the user-item affinity values associated with the campaign items of the number of the plurality of advertising campaigns;
assign the number of the plurality of advertising campaigns to a plurality of advertising campaign zones of a webpage of the website; and
transmit the assignment of the number of the plurality of advertising campaigns to the plurality of advertising campaign zones of the webpage to the server and to display to the user, wherein the server is configured to simultaneously display the plurality of advertising campaigns zones on the webpage to display to the user.

2. The system of claim 1 , wherein assigning the number of the plurality of advertising campaigns to the plurality of advertising campaign zones of the webpage comprises:
determining distance values between the advertising campaign zones of the webpage;
determining campaign similarity values between the number of the plurality of advertising campaigns; and
determining an assignment of at least a portion of the number of the plurality of advertising campaigns to the plurality of advertising campaign zones of the webpage based on the distance values and the campaign similarity values.

3. The system of claim 2 , wherein assigning the number of the plurality of advertising campaigns to the plurality of advertising campaign zones of the webpage comprises:
determining a first advertising campaign of the plurality of advertising campaigns with a highest user campaign value; and
assigning the first advertising campaign to a first advertising campaign zone of the plurality of advertising campaign zones of the webpage.

4. The system of claim 2 , wherein determining the assignment of the portion of the number of the plurality of advertising campaigns to the plurality of advertising campaign zones of the webpage based on the distance values and the campaign similarity values comprises:
determining a current advertising campaign zone of the plurality of advertising campaign zones;
determine a distance value between the current advertising campaign zone and any of the plurality of advertising campaign zones that have been assigned an advertising campaign of the plurality of advertising campaigns;
determine a campaign similarity value between each unassigned advertising campaign of the plurality of advertising campaigns and each assigned advertising campaign of the plurality of advertising campaigns; and
assign an unassigned advertising campaign of the plurality of advertising campaigns to the current advertising campaign zone based on the user campaign values, the distance values, and the campaign similarity values.

5. The system of claim 1 , wherein the computing device is configured to:
determine a ranking of the plurality of campaign items for each of the number of the plurality of advertising campaigns based on the user-item values, wherein transmitting the number of the plurality of advertising campaigns comprises transmitting the ranking of the plurality of campaign items for each of the number of the plurality of advertising campaigns.

6. The system of claim 1 , wherein the server is configured to transmit a request to the computing device when the user is accessing the webpage, and wherein the computing device is configured to transmit the number of the plurality of advertising campaigns to the server in response to the request.

7. The system of claim 1 , wherein determining, for each of the plurality of advertising campaigns, the user campaign value based on the campaign user data and the user attribute data comprises:
determining item attribute data identifying item attributes of at least one item previously purchased by the user; and
generating user embedding data by applying the neural network to the user attribute data and the item attribute data, wherein determining the user campaign value is based on the user embedding data.

8. The system of claim 1 , wherein the computing device is configured to:
for each of the plurality of advertising campaigns:
determine a number of the plurality of campaign items based on the user-item affinity values;
determine a revenue of the number of the plurality of campaign items;
determine a number of users that can still be assigned to the advertising campaign; and
adjust the user campaign value based on the revenue of the number of the plurality of campaign items and the number of users that can still be assigned to the advertising campaign, wherein assigning to the user the number of the plurality of advertising campaigns is based on the adjusted user campaign values.

9. A method comprising:
receiving campaign data identifying a plurality of advertising campaigns for a web site, each of the plurality of advertising campaigns iden

Assignees

• Walmart Apollo Llc

Inventors

• Arora Yokila
• Monemizadeh Morteza
• Mantha Aditya
• Guo Stephen Dean
• Achan Kannan

Classifications

• G06Q30/0242 Primary

  Determining effectiveness of advertisements · CPC title

• G06Q30/0271 Primary

  Personalized advertisement · CPC title

• G06Q30/0277

  Online advertisement · CPC title