Ontology-Based Natural Language Interpretation for Enterprise Software

Solution Overview

Problem

Conventional methods for interpreting natural language inputs in enterprise applications are inefficient and unreliable, often relying on computationally expensive statistical patterns that may not be available, leading to inaccurate and costly interpretations.

Innovation Solution

The use of a domain-specific ontology constructed from a database schema, combined with a Bayes network and contextual information, to predict the likelihood of ambiguous inputs referring to specific entities, enabling accurate and efficient interpretation of natural language inputs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If statistical pattern analysis is used to interpret natural language input, then the system can handle unstructured data, but the computational cost increases and reliability decreases

Engineering Contradiction:

Improveability to handle unstructured dataVSAvoidinterpretation accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent introduces an intermediary layer between unstructured natural language input and structured database queries. This intermediary consists of ontology construction from database schemas and probabilistic interpretation models that translate ambiguous user input into precise database operations, thereby maintaining reliability while handling unstructured data

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system performs preliminary action by pre-constructing ontologies from database schemas before actual user queries arrive. This pre-processing creates a structured framework that enables faster and more reliable interpretation of subsequent natural language inputs without requiring computational analysis at query time

Inventive Principle:

Principle #10Preliminary action

2Adaptability or versatility

If statistical pattern analysis is used to interpret natural language input, then the system can process diverse inputs, but the computational resources required increase

Engineering Contradiction:

Improveinput processing capabilityVSAvoidcomputational cost

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary action by pre-constructing ontologies from database schemas before actual user queries arrive. This pre-processing creates a structured framework that enables faster and more reliable interpretation of subsequent natural language inputs without requiring computational analysis at query time

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter of interpretation from statistical pattern matching to ontology-based semantic matching. This parameter change transforms the computational approach from resource-intensive statistical analysis to more efficient structured query evaluation based on pre-defined semantic relationships

Inventive Principle:

Principle #35Parameter changes

3Ease of operation

If ambiguous input interpretation is made more lenient, then user experience improves, but interpretation accuracy may decrease

Engineering Contradiction:

Improveuser input simplicityVSAvoidinput interpretation accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The system implements feedback through probabilistic interpretation that calculates confidence scores for each possible meaning of ambiguous input. When confidence is high, the system proceeds with interpretation; when confidence is low, it can request clarification or provide multiple interpretations, thereby maintaining both ease of use and accuracy

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces an intermediary layer between unstructured natural language input and structured database queries. This intermediary consists of ontology construction from database schemas and probabilistic interpretation models that translate ambiguous user input into precise database operations, thereby maintaining reliability while handling unstructured data

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11347783B2Implementing a software action based on machine interpretation of a language input

Publication Date: 2022.05.31 ORACLE INT CORP

AI summary

A schema-ontology is automatically constructed with reference to implicit sematic relationships of a database schema. An estimation of the meaning of the language input is determined based on the language input and the schema-ontology. The machine interpretation of the language input is generated based on the meaning and based on the estimation of the meaning including an estimation of an ambiguity of portions of the language input. A software action that is responsive to the machine interpretation of the language input is selected. The software action is implemented based on the machine interpretation of the language input.