Meaning Representation Query Interpretation for Transparent NL2SQL

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing natural language to logical form conversion systems, such as NL2SQL, require large amounts of labeled data for deep learning and struggle to generalize across domains, leading to inefficiencies in translating user queries to database queries, and lack clarity in user query interpretation.

Innovation Solution

A machine-learning based system converts natural language utterances to logical form queries using a Meaning Resource Language (MRL) structure, generating readable natural language interpretations of the logical form queries, and executes system queries to provide transparent user understanding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If deep learning approaches are used to convert natural language queries to SQL queries, then the system can generalize across different domains, but large amounts of labeled training data are required

Engineering Contradiction:

Improvegeneralization capabilityVSAvoidlabeled training data

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent introduces Meaning Representation Language (MRL) as an intermediary layer between natural language queries and SQL queries. The semantic parser converts natural language to MRL (a structured logical form), which then serves as intermediate representation that can be more easily translated to SQL. This intermediary structure reduces the need for extensive labeled training data while maintaining generalization capability across domains.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Quantity of substance

If rule-based systems are used to translate natural language sentences to SQL queries, then the system is domain-specific and inelastic, but it requires less training data

Engineering Contradiction:

Improvetraining dataVSAvoiddomain generalization

Core Design Contradiction:

Quantity of substanceVSAdaptability or versatility

Solution Approach 1:

The translation process is segmented into distinct stages: natural language understanding, semantic parsing to MRL structure, and SQL generation. Each stage handles specific aspects of the translation, allowing the system to be more adaptable to different domains while maintaining efficiency. The MRL structure acts as a domain-agnostic intermediate representation that facilitates generalization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system employs dynamic element selection during semantic parsing, where the parser can adaptively choose different parsing strategies and structural representations based on the input query characteristics. This dynamic approach allows the system to handle diverse query types and domains effectively without requiring rigid domain-specific rules.

Inventive Principle:

Principle #15Dynamics

3Ease of operation

If statistical-based slot-filling systems are used to parse natural language sentences, then the system can handle structured queries, but it is limited to sentences that can be represented as parse trees

Engineering Contradiction:

Improvestructured query handlingVSAvoidsentence structure flexibility

Core Design Contradiction:

Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent changes the structural parameters of the intermediate representation from rigid parse tree structures to flexible MRL structures with configurable elements and relationships. This allows the system to handle a broader range of sentence structures while maintaining the ability to process structured queries effectively.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12579137B2Output interpretation for a meaning representation language system

Publication Date: 2026.03.17 ORACLE INT CORP

AI summary

The present disclosure is related to techniques for converting a natural language utterance to a logical form query and deriving a natural language interpretation of the logical form query. The techniques include accessing a Meaning Resource Language (MRL) query and converting the MRL query into a MRL structure including logical form statements. The converting includes extracting operations and associated attributes from the MRL query and generating the logical form statements from the operations and associated attributes. The techniques further include translating each of the logical form statements into a natural language expression based on a grammar data structure that includes a set of rules for translating logical form statements into corresponding natural language expressions, combining the natural language expressions into a single natural language expression, and providing the single natural language expression as an interpretation of the natural language utterance.