AI Query Profile Selection for Secure Natural Language SQL
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database systems face challenges in efficiently processing natural language queries due to ambiguity and security risks, as they struggle to accurately interpret user intent and enforce role-based access controls, leading to potential data exposure and inefficient query execution.
Innovation Solution
A computer-implemented method that utilizes a large language model to generate SQL requests, appending security predicates to enforce access constraints, and visualizes data based on categorized natural language inputs, ensuring secure and efficient database access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a user is given the ability to access all data of an organization or manipulate data outside one of the organization's application environments, then the user can perform comprehensive data analysis and visualization, but large security risks are introduced for the organization
Solution Approach 1:
The patent introduces an intermediary layer between the user and the database that automatically generates and enforces security predicates. This mediator translates user intent from natural language queries into secure SQL commands with appropriate access constraints, allowing comprehensive data analysis while maintaining security boundaries without requiring users to manually configure access controls.
Solution Approach 2:
The system performs preliminary action by pre-defining security predicates and access constraints before data queries are executed. The intermediary layer prepares and validates security rules in advance, ensuring that only authorized data access occurs while still enabling users to perform comprehensive analysis within their permission boundaries.
2Reliability
If role-based access controls are enforced at the database level, then data security is maintained, but the specific boundaries of access control for individual users are difficult to determine and applications cannot predict what data users will see
Solution Approach 1:
The system implements feedback by having the intermediary layer continuously monitor and analyze user queries against defined security predicates. It provides real-time determination of access boundaries by evaluating user intent and comparing it with security rules, then returns predictable information about what data the user will actually see based on their role and the current query context.
Solution Approach 2:
The patent replaces the traditional mechanical approach of manually configuring and tracking role-based access controls with an intelligent system that uses natural language processing and automated reasoning. The intermediary layer substitutes complex access control determination mechanics with AI-driven analysis that automatically understands user intent and applies appropriate security boundaries.
3Device complexity
If natural language queries are processed without categorization, then query processing is simpler, but accuracy in interpreting user intent and selecting appropriate database structures decreases
Solution Approach 1:
The system applies segmentation by dividing the natural language query processing into distinct categorical stages: query categorization, profile selection, security predicate determination, and SQL generation. This segmentation improves accuracy by handling each aspect of user intent separately while maintaining overall process simplicity through automated transitions between stages.
Data Source
AI summary
A database query processing method includes receiving a natural language request for information contained within a database, categorizing the natural language request in a particular category from a plurality of categories, and selecting a large language model and one or more database structures for which a SQL request is to be directed based in part on the particular category. The selected large language model is prompted to generate a SQL request directed to the selected set of one ore more database structures and the received SQL request is validated for the particular category. The validated particular SQL request is used to access data within the database and one or more visualizations is caused to be displayed based at least in part on a result of the validated particular SQL request.