Systems and methods for relation labelling pipeline

Patent metadata

Field	Value
Publication number	US-2025245426-A1
Application number	US-202519030549-A
Country	US
Kind code	A1
Filing date	Jan 17, 2025
Priority date	Jan 31, 2024
Publication date	Jul 31, 2025
Grant date	—

Abstract

Systems and methods for generating an interface including recommended elements selected using generated element type relation labels are disclosed. An interface generation request including at least one element type is received and a set of recommended elements is generated based on element type relations between the at least one element type and additional element types associated with a network interface. The element type relations are generated by at least one large language model and at least one optimal relation generation prompt. An interface including the set of recommended elements is generated.

First claim

What is claimed is:

1 . A system, comprising:
a processor; and a non-transitory memory storing instructions that, when executed, cause the processor to:
receive an interface generation request including at least one element type;
generate a set of recommended elements based on element type relations between the at least one element type and one or more additional element types associated with a network interface, wherein the element type relations are generated by at least one large language model and at least one optimal relation generation prompt; and
generate an interface including the set of recommended elements.

2 . The system of claim 1 , wherein the set of recommended elements are generated by one of a similar item recommendation process or a complementary item recommendation process.

3 . The system of claim 1 , wherein the element type relations are generated based on an anchor item.

4 . The system of claim 1 , wherein the element type relations comprise doublets including a first type and a second type.

5 . The system of claim 1 , wherein the at least one large language model receives a prompt generated by a template completion process.

6 . The system of claim 5 , wherein the prompt comprises relational label definitions.

7 . The system of claim 6 , wherein the relational label definition are unidirectional.

8 . A computer-implemented method, comprising:
receiving an interface generation request including at least one element type; generating a set of recommended elements based on element type relations between the at least one element type and additional element types associated with a network interface, wherein the element type relations are generated by at least one large language model and at least one optimal relation generation prompt; and generating an interface including the set of recommended elements.

9 . The computer-implemented method of claim 8 , wherein the set of recommended elements are generated by one of a similar item recommendation process or a complementary item recommendation process.

10 . The computer-implemented method of claim 8 , wherein the element type relations are generated based on an anchor item.

11 . The computer-implemented method of claim 8 , wherein the element type relations comprise doublets including a first type and a second type.

12 . The computer-implemented method of claim 8 , wherein the at least one large language model receives a prompt generated by a template completion process.

13 . The computer-implemented method of claim 12 , wherein the prompt comprises relational label definitions.

14 . The computer-implemented method of claim 13 , wherein the relational label definition are unidirectional.

15 . A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by at least one processor, cause at least one device to perform operations comprising:
receiving an interface generation request including at least one element type; generating a set of recommended elements based on element type relations between the at least one element type and additional element types associated with a network interface, wherein the element type relations are generated by at least one large language model and at least one optimal relation generation prompt; and generating an interface including the set of recommended elements.

16 . The non-transitory computer readable medium of claim 15 , wherein the set of recommended elements are generated by one of a similar item recommendation process or a complementary item recommendation process.

17 . The non-transitory computer readable medium of claim 15 , wherein the element type relations are generated based on an anchor item.

18 . The non-transitory computer readable medium of claim 15 , wherein the element type relations comprise doublets including a first type and a second type.

19 . The non-transitory computer readable medium of claim 15 , wherein the at least one large language model receives a prompt generated by a template completion process.

20 . The non-transitory computer readable medium of claim 15 , wherein the prompt comprises unidirectional relational label definitions.

Assignees

• Walmart Apollo Llc

Inventors

• Chen Jiao
• Ma Luyi
• Li Xiaohan
• Thakurdesai Nikhil Shripad
• Xu Jianpeng
• Cho Hyun Duk
• Nag Kaushiki
• Korpeoglu Evren
• Kumar Sushant
• Achan Kannan

Classifications

• G06F16/383

  using metadata automatically derived from the content · CPC title

• G06Q30/0631

  Recommending goods or services · CPC title

• G06F40/186 Primary

  Templates · CPC title