Concept Graph Schema Linking for Efficient Natural Language Queries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems face inflexibility and inefficiency in executing structured database queries, often requiring rigid SQL formats and excessive computational resources due to over-selection of data, especially in large databases with many tables.
Innovation Solution
A schema linking system generates a concept graph linking database tables to concepts based on their content, using a large language model to convert natural language queries into structured queries that target only relevant tables, reducing processing overhead and resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional systems require rigid SQL format for structured database queries, then query execution follows a defined syntax, but flexibility of operation deteriorates
Solution Approach 1:
The patent introduces a natural language processing intermediary layer between the user and the database query system. This intermediary translates flexible natural language queries into structured SQL queries, allowing users to operate the system in their own language while maintaining the structured execution required by the database. The system acts as a mediator that converts between the flexible input format and the rigid execution format.
2Reliability
If conventional systems search large amounts of data across many different tables in databases, then comprehensive information can be retrieved, but excessive computational resources are required
Solution Approach 1:
The patent extracts and identifies only the relevant tables needed to answer a specific natural language query before executing the search. Instead of searching across all tables in the database, the system analyzes the query, identifies the specific tables containing relevant information, and limits the search scope to those extracted tables. This significantly reduces computational resources while maintaining comprehensive information retrieval within the relevant scope.
3Productivity
If conventional systems make large quantities of calls to machine learning models for each query, then query processing can be performed, but resource costs increase
Solution Approach 1:
The patent performs preliminary analysis of the natural language query to identify relevant tables and construct an optimized query plan before executing the actual search. By preprocessing the query and determining the necessary tables in advance, the system avoids making multiple redundant calls to machine learning models during query execution. This preliminary action reduces resource costs while maintaining full query processing capability.
Data Source
AI summary
The present disclosure relates to systems, non-transitory computer-readable media, and methods for using a concept graph to select relevant tables relevant for querying a database. In particular, in some embodiments, the disclosed systems generate concept tags for tables in a database schema based on content of the tables corresponding to concepts in a list of concepts. Additionally, the disclosed systems generate a concept graph comprising hyper edges linking the tables to the concepts according to the concept tags. The disclosed systems determine, from the tables in the database schema, a set of tables relevant to a natural language query comprising an indicated concept by extracting the set of tables from one or more hyper edges corresponding to the indicated concept from the concept graph. The disclosed systems also generate, utilizing a large language model, a response for the natural language query from the set of relevant tables.