Data query method and apparatus based on large model, electronic device, and storage medium

Patent metadata

Field	Value
Publication number	US-12505092-B2
Application number	US-202418974155-A
Country	US
Kind code	B2
Filing date	Dec 9, 2024
Priority date	Sep 18, 2024
Publication date	Dec 23, 2025
Grant date	Dec 23, 2025

Abstract

Data query method and apparatus based on large model, an electronic device, and a storage medium are disclosed, which relates to the field of artificial intelligence, specifically in natural language processing, deep learning, and large model technologies, applicable to scenarios such as dialogue systems and information retrieval. The method includes: performing entity recognition on a query to obtain the target entity in the query; obtaining a first related content associated with the target entity from internal information, and performing data analysis on the first related content using a large language model (LLM) to obtain a data analysis result; obtaining a second related content associated with the target entity from external information, and performing data generation on the second related content using the LLM to obtain a data generation result; obtaining a query result corresponding to the query based on the data analysis result and the data generation result.

First claim

What is claimed is:

1 . A computer-implemented method for data query based on large model, comprising:
performing entity recognition on a query input by a user in natural language to obtain a target entity in the query, where the target entity comprises one or more of geographic information, time information, and domain information; converting the target entity into an entity vector; obtaining a target content that matches the entity vector in a database from internal information so that the target entity in the query is correctly mapped to the corresponding field in the database; obtaining impact factors related to the domain information of the target entity based on a knowledge graph and analyzing the trends of the impact factors related to the geographic information and the time information of the target entity to obtain a relevant content of the target content; using the target content and the relevant content as first related content, and performing data analysis on the first related content using a large language model (LLM) to obtain a data analysis result; obtaining a second related content associated with the target entity from external information, and performing data generation on the second related content using the LLM to obtain a data generation result; and obtaining a query result corresponding to the query based on the data analysis result and the data generation result.

2 . The method of claim 1 , wherein performing data analysis on the first related content using the LLM to obtain the data analysis result comprises:
analyzing the first related content using the LLM to obtain an analysis result; calibrating the query using the LLM based on the analysis result to obtain a calibration result; and summarizing the analysis result and the calibration result using the LLM to obtain the data analysis result.

3 . The method of claim 1 , wherein obtaining the second related content associated with the target entity from the external information comprises:
obtaining external information that matches with the time of the query based on a publication time of the external information; and obtaining the second related content associated with the target entity from the external information that matches with the time of the query.

4 . The method of claim 1 , wherein performing data generation on the second related content using the LLM to obtain the data generation result comprises:
filtering the second related content using the LLM to obtain a filtering result; and performing generation on the data filtering result using the LLM based on an information source corresponding to the filtering result to obtain the data generation result.

5 . The method of claim 1 , wherein performing entity recognition on the query to obtain the target entity in the query comprises:
performing entity recognition on the query based on the LLM to obtain the target entity.

6 . An electronic device, comprising:
at least one processor; and a memory communicatively connected with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform a method for data query based on large model, wherein the method for data query based on large model comprises: performing entity recognition on a query input by a user in natural language to obtain a target entity in the query, where the target entity comprises one or more of geographic information, time information, and domain information; converting the target entity into an entity vector; obtaining a target content that matches the entity vector in a database from internal information so that the target entity in the query is correctly mapped to the corresponding field in the database; obtaining impact factors related to the domain information of the target entity based on a knowledge graph and analyzing the trends of the impact factors related to the geographic information and the time information of the target entity to obtain a relevant content of the target content; using the target content and the relevant content as first related content, and performing data analysis on the first related content using a large language model (LLM) to obtain a data analysis result; obtaining a second related content associated with the target entity from external information, and performing data generation on the second related content using the LLM to obtain a data generation result; and obtaining a query result corresponding to the query based on the data analysis result and the data generation result.

7 . The electronic device of claim 6 , wherein performing data analysis on the first related content using the LLM to obtain the data analysis result comprises:
analyzing the first related content using the LLM to obtain an analysis result; calibrating the query using the LLM based on the analysis result to obtain a calibration result; and summarizing the analysis result and the calibration result using the LLM to obtain the data analysis result.

8 . The electronic device of claim 6 , wherein obtaining the second related content associated with the target entity from the external information comprises:
obtaining external information that matches with the time of the query based on a publication time of the external information; and obtaining the second related content associated with the target entity from the external information that matches with the time of the query.

9 . The electronic device of claim 6 , wherein performing data generation on the second related content using the LLM to obtain the data generation result comprises:
filtering the second related content using the LLM to obtain a filtering result; and performing generation on the data filtering result using the LLM based on an information source corresponding to the filtering result to obtain the data generation result.

10 . The electronic device of claim 6 , wherein performing entity recognition on the query to obtain the target entity in the query comprises:
performing entity recognition on the query based on the LLM to obtain the target entity.

11 . A non-transitory computer readable storage medium with computer instructions stored thereon, wherein the computer instructions are used for causing a method for data query based on large model, wherein the method for data query based on large model comprises:
performing entity recognition on a query input by a user in natural language to obtain a target entity in the query, where the target entity comprises one or more of geographic information, time information, and domain information; converting the target entity into an entity vector; obtaining a target content that matches the entity vector in a database from internal information so that the target entity in the query is correctly mapped to the corresponding field in the database; obtaining impact factors related to the domain information of the target entity based on a knowledge graph and analyzing the trends of the impact factors related to the geographic information and the time information of the target entity to obtain a relevant content of the target content; using the target content and the relevant content as first related content, and performing data analysis on the first related content using a large language model (LLM) to obtain a data analysis result; obtaining a second related content associated with the target entity from external information, and performing data generation on the second related content using the LLM to obtain a data generation result; and obtaining a query result corresponding to

Assignees

• Beijing Baidu Netcom Sci & Tech Co Ltd

Inventors

• Xiao Yixiong
• Zhou Jingbo

Classifications

• G06F40/295

  Named entity recognition · CPC title

• G06F16/24522

  Translation of natural language queries to structured queries · CPC title

• G06F16/2282

  Tablespace storage structures; Management thereof · CPC title

• G06F16/243

  Natural language query formulation · CPC title

• G06F16/367

  Ontology · CPC title