Facilitating user selection using trend-based joint embeddings

Patent metadata

Field	Value
Publication number	US-11928719-B2
Application number	US-202117543239-A
Country	US
Kind code	B2
Filing date	Dec 6, 2021
Priority date	Dec 6, 2021
Publication date	Mar 12, 2024
Grant date	Mar 12, 2024

Abstract

Methods, systems, and computer program products for facilitating user selection using trend-based joint embeddings are provided herein. A method includes obtaining a selection of an item in an online catalog; determining a compatible item of the plurality of items at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained on historical data associated with the item in the online catalog and fine-tuned based on current trend data from multiple data sources; receiving feedback in response to outputting the at least one compatible item; identifying one or more attributes related to the at least one compatible item based on the feedback; and using the trend-based machine learning model to determine at least one additional compatible item based on the one or more attributes.

First claim

What is claimed is:

1. A computer-implemented method, the method comprising:
obtaining a selection from a user of at least one of a plurality of items in an online catalog;
determining at least one compatible item of the plurality of items in the online catalog at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained to learn joint embeddings between items such that, for a given embedding provided as input, the trend-based machine learning model outputs a similar embedding of an item, and wherein the trend-based machine learning model is trained on historical data associated with at least a portion of the plurality of items in the online catalog and fine-tuned based on current trend data, from multiple data sources, in multiple different modalities;
receiving feedback from the user in response to outputting the at least one compatible item;
identifying one or more attributes related to the at least one compatible item based at least in part on the feedback; and
using the trend-based machine learning model to determine at least one additional compatible item in the online catalog based at least in part on the one or more attributes;
wherein the method is carried out by at least one computing device.

2. The computer-implemented method of claim 1 , wherein the selection from the user and the feedback from the user are provided in a natural language format.

3. The computer-implemented method of claim 1 , wherein the selection of the user is processed using a natural language decoder of a software agent that determines one or more keywords of the user selection, and wherein the providing comprises transforming the one or more keywords into a format that is processable by the trend-based machine learning model.

4. The computer-implemented method of claim 1 , wherein the similiar embedding of the item output by the trend-based machine learning model is in a different category than the item corresponding to the given embedding.

5. The computer-implemented method of claim 1 , wherein the current trend data comprises at least two of:
data from one or more online social media sites;
data from one or more subject matter experts;
sales data over a specified time-period; and
co-occurrence of items with a plurality of users.

6. The computer-implemented method of claim 1 , wherein the at least one previously selected item comprises an item identified based on at least one of a profile of the user or a purchase history of the user.

7. The computer-implemented method of claim 1 , wherein the identifying comprises:
applying a counterfactual query process that perturbs at least one attribute of the selected item based on the feedback to increase a compatibility score with the at least one previously selected item.

8. The computer-implemented method of claim 7 , wherein the compatibility score is calculated based on a cosine distance between embeddings of counterfactual queries generated by the counterfactual process and embeddings of the at least one previously selected item.

9. The computer-implemented method of claim 1 , wherein the at least one item comprises a product having a first type and the at least one compatible product comprises another product having a different, second type that is related to the first type.

10. The computer-implemented method of claim 9 , wherein each of the first type and the second type is a type of apparel.

11. The computer-implemented method of claim 1 , wherein software is provided as a service in a cloud environment for implementing at least a part of the trend-based machine learning model.

12. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:
obtain a selection from a user of at least one of a plurality of items in an online catalog;
determine at least one compatible item of the plurality of items in the online catalog at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained to learn joint embeddings between items such that, for a given embedding provided as input, the trend-based machine learning model outputs a similar embedding of an item, and wherein the trend-based machine learning model is trained on historical data associated with at least a portion of the plurality of items in the online catalog and fine-tuned based on current trend data, from multiple data sources, in multiple different modalities;
receive feedback from the user in response to outputting the at least one compatible item;
identify one or more attributes related to the at least one compatible item based at least in part on the feedback; and
use the trend-based machine learning model to determine at least one additional compatible item in the online catalog based at least in part on the one or more attributes.

13. The computer program product of claim 12 , wherein the selection from the user and the feedback from the user are provided in a natural language format.

14. The computer program product of claim 12 , wherein the selection of the user is processed using a natural language decoder of a software agent that determines one or more keywords of the user selection, and wherein the providing comprises transforming the one or more keywords into a format that is processable by the trend-based machine learning model.

15. The computer program product of claim 12 , wherein the similar embedding of the item output by the trend-based machine learning model is in a different category than the item corresponding to the given embedding.

16. The computer program product of claim 12 , wherein the current trend data comprises at least two of:
data from one or more online social media sites;
data from one or more subject matter experts;
sales data over a specified time-period; and
co-occurrence of items with a plurality of users.

17. The computer program product of claim 12 , wherein the at least one previously selected item comprises an item identified based on at least one of a profile of the user or a purchase history of the user.

18. The computer program product of claim 12 , wherein the identifying comprises:
applying a counterfactual query process that perturbs at least one attribute of the selected item based on the feedback to increase a compatibility score with the at least one previously selected item.

19. The computer program product of claim 12 , wherein the at least one item comprises a product having a first type and the at least one compatible product comprises another product having a different, second type that is related to the first type.

20. A system comprising:
a memory configured to store program instructions;
a processor operatively coupled to the memory to execute the program instructions to:
obtain a selection from a user of at least one of a plurality of items in an online catalog;
determine at least one compatible item of the plurality of items in the online catalog at least in part by providing the selected at least one item and at least one previously selected item corresponding to the user to a trend-based machine learning model, wherein the trend-based machine learning model is trained t

Assignees

• IBM

Inventors

• Dwivedi Satyam
• Ekambaram Vijay
• Manglik Kushagra
• Aggarwal Nupur
• Raykar Vikas C

Classifications

• G06Q10/44

  Identification of trends within social networks, e.g. identification of trending topics · CPC title

• G06Q30/0631 Primary

  Recommending goods or services · CPC title

• G06F40/40

  Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) · CPC title

• G06Q30/0629 Primary

  by pre-processing results, e.g. ranking or ordering results · CPC title

• H04L51/046

  Interoperability with other network applications or services · CPC title