Text-Based Query Translation for Relational Database Access

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Relational databases are difficult for non-expert users to query effectively due to the complexity of Structured Query Language (SQL), making it impractical for routine inquiries in settings like physical shopping facilities without constant expert availability or extensive user training.

Innovation Solution

A control circuit that translates text-based queries into Structured Query Language (SQL) queries, allowing users to access relational databases using a simple, browser-like interface, without revealing the SQL queries to the user, and accommodating Boolean operators and multiple databases.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If SQL is used to query relational databases, then query precision and accuracy are improved, but ease of operation deteriorates

Engineering Contradiction:
Improvequery accuracyVSAvoiduser-friendliness
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent introduces a natural language processing intermediary layer that translates user-friendly text queries into SQL queries. This intermediary system includes a query parser, semantic analyzer, and SQL generator that bridge the gap between simple user input and complex database queries, allowing non-experts to access relational databases without learning SQL syntax while maintaining query accuracy

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If SQL expertise is required for database queries, then query reliability is improved, but ease of operation deteriorates

Engineering Contradiction:
Improvequery reliabilityVSAvoiduser accessibility
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The system enables self-service database querying by automatically translating natural language inputs into executable SQL queries. The patent implements automated query construction, validation, and execution without requiring human SQL expertise. The system includes error detection, query optimization, and result verification mechanisms that maintain reliability while eliminating the need for specialized knowledge

Inventive Principle:
Principle #25Self-service

3Ease of operation

If comprehensive user training is provided, then ease of operation is improved, but loss of time increases

Engineering Contradiction:
Improveuser proficiencyVSAvoidtraining time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent pre-compiles and stores natural language query templates and their corresponding SQL translations in a query dictionary. Common business queries are预先 prepared with optimized SQL formulations, allowing the system to quickly match and execute queries without requiring users to learn complex syntax or spend time on training. The system includes template matching, fuzzy searching, and automatic query refinement capabilities

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS10445808B2Apparatus to query a relational database using text-based queries
Publication Date: 2019.10.15 WALMART APOLLO LLC
  • US10445808B2 patent drawing
  • US10445808B2 patent drawing
  • US10445808B2 patent drawing

AI summary

A control circuit has access to both a relational database comprised of structured data and a text-based index of reciprocal queries for the relational database. The control circuit, upon receiving a text-based query via a user input, applies the text-based query to the text-based index of reciprocal queries to thereby obtain one or more reciprocal queries that the control circuit uses to query the relational database for structured data content responsive to the text-based query. By one approach the text-based index of reciprocal queries comprises a text-based index of SQL reciprocal queries. By one approach the text-based index of reciprocal queries comprises at least one virtual document.