Methods and apparatus for generating item recommendations based on real-time inference of machine learning models

Patent metadata

Field	Value
Publication number	US-11645559-B2
Application number	US-202017017114-A
Country	US
Kind code	B2
Filing date	Sep 10, 2020
Priority date	Sep 10, 2020
Publication date	May 9, 2023
Grant date	May 9, 2023

Abstract

This application relates to apparatus and methods for providing recommended items to advertise. In some examples, a computing device determines a plurality of first values for a corresponding plurality of first items based on the user's engagement with each of the first items. The computing device may then determine a subset of the plurality of first items based on the first values. The computing device may receive a search request and determine a plurality of second values for a plurality of second items based on the search request. The computing device may determine a plurality of third values for the subset of items based on the plurality of second values for the plurality of second items and the user's engagement with each of the subset of items. The computing device may determine the recommended items based on the plurality of second values and the plurality of third values.

First claim

What is claimed is:

1. A system comprising:
a processing unit configured to:
receive user session data identifying website activities of a user from a web server;
generate feature data based on at least the user session data for the user;
provide a first portion of the feature data to a second processing unit;
assign a first machine learning model to the second processing unit for execution to generate first output data based on the first portion of the feature data;
provide a second portion of the feature data to a third processing unit;
assign a second machine learning model to the third processing unit for execution to generate second output data based on the second portion of the feature data;
receive the first output data from the second processing unit;
receive the second output data from the third processing unit;
generate item recommendations based on the first output data and the second output data; and
transmit the item recommendations to the web server.

2. The system of claim 1 , wherein:
the processing unit has a first type of architecture;
the second processing unit has a second type of architecture; and
the third processing unit has the second type of architecture.

3. The system of claim 2 , wherein the second processing unit and the third processing unit are configured to execute at least a portion of the first machine learning model and at least a portion of the second machine learning model, respectively, in parallel.

4. The system of claim 2 , wherein each of the first machine learning model and the second machine learning model are generative models.

5. The system of claim 1 , wherein the item recommendations are generated based on:
executing a merging algorithm based on the first output data and the second output data to generate third output data; and
generating the item recommendations based on the third output data.

6. The system of claim 1 , wherein the processing unit is a central processing unit (CPU), and each of the at least one different processing unit is a graphical processing unit (GPU).

7. The system of claim 1 , wherein the processing unit is configured to provide the first output data and the second output data to the second processing unit for execution of a third machine learning model to generate third output data, and wherein the item recommendations are generated based on the third output data.

8. The system of claim 7 , wherein
the processing unit is configured to
assign the third machine learning model to the second processing unit when the first output data and the second output data have been generated.

9. The system of claim 8 , wherein the processing unit is configured to execute the assignments based on configuration data stored in a database, wherein the configuration data identifies input requirements for each of the first machine learning model, the second machine learning model, and the third machine learning model.

10. A method by a processing unit comprising:
receiving user session data identifying website activities of a user from a web server;
generating feature data based on at least the user session data for the user;
providing a first portion of the feature data to a second processing unit;
assigning a first machine learning model to the second processing unit for execution to generate first output data based on the first portion of the feature data;
providing a second portion of the feature data to a third processing unit;
assigning a second machine learning model to the third processing unit for execution to generate second output data based on the second portion of the feature data;
receiving the first output data from the second processing unit;
receiving the second output data from the third processing unit;
generating item recommendations based on the first output data and the second output data; and
transmitting the item recommendations to the web server.

11. The method of claim 10 , wherein:
the first machine learning model and the second machine learning model are assigned based on configuration data stored in a database; and
the configuration data identifies input requirements for each of the first machine learning model and the second machine learning model.

12. The method of claim 11 , further comprising causing the second processing unit and the third processing unit to execute at least a portion of the first machine learning model and at least a portion of the second machine learning model in parallel.

13. The method of claim 10 , wherein generating the item recommendations comprises:
executing a merging algorithm based on the first output data and the second output data to generate third output data; and
generating the item recommendations based on the third output data.

14. The method of claim 10 , wherein the processing unit is a central processing unit (CPU), and each of the at least one different processing unit is a graphical processing unit (GPU).

15. The method of claim 10 , further comprising providing the first output data and the second output data to the second processing unit for execution of a third machine learning model to generate third output data, and wherein generating the item recommendations is based on the third output data.

16. A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by at least one processor, cause a device to perform operations comprising:
receiving user session data identifying website activities of a user from a web server;
generating feature data based on at least the user session data for the user;
providing a first portion of the feature data to a second processing unit;
assigning a first machine learning model to the second processing unit for execution to generate first output data based on the first portion of the feature data;
providing a second portion of the feature data to a third processing unit;
assigning a second machine learning model to the third processing unit for execution to generate second output data based on the second portion of the feature data;
receiving the first output data from the second processing unit;
receiving the second output data from the third processing unit;
generating item recommendations based on the first output data and the second output data; and
transmitting the item recommendations to the web server.

17. The non-transitory computer readable medium of claim 16 , wherein:
the first machine learning model and the second machine learning model are assigned based on configuration data stored in a database; and
the configuration data identifies input requirements for each of the first machine learning model and the second machine learning model.

18. The non-transitory computer readable medium of claim 17 , wherein the instructions, when executed by the at least one processor, further cause the device to perform operations comprising causing the second processing unit and the third processing unit to execute at least a portion of the first machine learning model and at least a portion of the second machine learning model in parallel.

19. The non-transitory computer readable medium of claim 16 , wherein the instructions, when executed by the at least one processor, further cause the device to perform operations comprising:
executing a merging algorithm based on the first output data and the second output data to generate third output data; and
generating the item recommendations based on the third output data.

20. The non-transitory

Assignees

• Walmart Apollo Llc

Inventors

• Yalla Surya
• Nag Kaushiki
• Manoharan Shirpaa
• Achan Kannan

Classifications

• G06N5/04 Primary

  Inference or reasoning models · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title

• G06N20/00

  Machine learning · CPC title