Rule-Based NLIDB Refinement via Deep Learning Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual and time-consuming process of generating database ontologies for rule-based natural language interfaces, which requires deep knowledge of the database structure and schema, is inefficient and not scalable for large or complex systems.
Innovation Solution
A method that uses a parallel deep learning model in conjunction with a rule-based natural language interface to process natural language queries, where both models generate explanations and are validated by users, allowing for automatic refinement of rules and updates to improve the accuracy and efficiency of the interface without extensive manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a rule-based natural language interface is used to convert natural language queries to structured database queries, then the conversion is fast and simple, but the system requires extensive manual knowledge and expertise in database structure and schema
Solution Approach 1:
The system automatically refines its own rules by using the deep learning model to suggest rule refinements based on query examples, which are then validated and integrated into the rule-based system without requiring external manual intervention for each rule creation
Solution Approach 2:
The patent replaces the manual mechanical process of rule creation requiring expert knowledge with an automated deep learning model that learns patterns from data and generates rule refinements automatically
2Manufacturing precision
If manual rule generation is used for database ontology creation, then the rules can be precisely tailored to database structure, but the process is time-consuming and not scalable
Solution Approach 1:
The deep learning model is pre-trained on database schemas and query patterns, enabling it to quickly generate informed rule refinements without requiring manual rule creation from scratch for each new database or query type
Solution Approach 2:
The system automatically refines its own rules by using the deep learning model to suggest rule refinements based on query examples, which are then validated and integrated into the rule-based system without requiring external manual intervention for each rule creation
3Adaptability or versatility
If a deep learning model is used to process natural language queries, then the system can learn and adapt to different query patterns, but the model requires extensive training data and computational resources
Solution Approach 1:
The patent merges a deep learning model with a rule-based system, where the deep learning model handles pattern recognition and adaptation while the rule-based system provides structured query generation, combining the strengths of both approaches
Solution Approach 2:
The rule-based system acts as an intermediary between the deep learning model and the database, translating the model's learned patterns into structured queries while the deep learning model serves as an intermediary between natural language input and rule-based processing
Data Source
AI summary
A computer-implemented method, a computer system and a computer program product create rules for a rule-based natural language interface for databases (NLIDB). The method may include receiving a natural language query from a user. The method may also include generating a first explanation for the natural language query using a deep learning model and a second explanation for the natural language query using the rule-based NLIDB and validating whether the first and second explanations correctly represent the natural language query. The method may further include identifying the database value in the first explanation in response to the first explanation correctly representing the natural language query and the second explanation not correctly representing the natural language query. Lastly, the method may include creating a rule in a table for the rule-based natural language interface for databases that associates the database value with the original word of the natural language query.