Dialog Application Service Using NLU Model Generation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The development and deployment of dialog-driven applications for various domains and devices are complex, requiring expertise in programming dialog flows, automated speech recognition, and natural language processing, and often involve platform-specific details, making it difficult for developers to create applications that can handle multi-step interactions and integrate with different resources efficiently.

Innovation Solution

A network-accessible service that utilizes natural language understanding (NLU) components and programmatic interfaces allows developers to specify dialog steps and intents without writing source code, generating executable applications that can manage multi-step interactions across various devices and platforms, using automated speech recognition and natural language processing algorithms to interpret user inputs and integrate with diverse resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If developers manually program dialog flows, speech recognition, and natural language processing, then application functionality and customization are improved, but development complexity and time consumption increase significantly

Engineering Contradiction:
Improveapplication functionalityVSAvoiddevelopment complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces a service as an intermediary between developers and the complex NLU system. This service provides pre-built NLU components and automated generation tools that mediates the complexity, allowing developers to create customized applications without manually programming speech recognition and natural language processing algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables self-service through automated generation of application portions based on example inputs. The NLU model automatically generates dialog flows, speech recognition patterns, and natural language processing logic without requiring manual programming, reducing development complexity while maintaining functionality.

Inventive Principle:
Principle #25Self-service

2Reliability

If platform-specific details are handled manually, then application performance and optimization are improved, but ease of deployment and portability worsen

Engineering Contradiction:
Improveapplication performanceVSAvoidease of deployment
Core Design Contradiction:
ReliabilityVSEase of manufacture

Solution Approach 1:

The service provides a universal platform that handles multiple platforms and devices through a common interface. The generated applications can be deployed across different platforms without manual reconfiguration, as the service abstracts platform-specific details while maintaining optimized performance through its standardized NLU components.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Ease of operation

If comprehensive natural language understanding is implemented, then user interaction quality is improved, but computational resources and processing time increase

Engineering Contradiction:
Improveuser interaction qualityVSAvoidcomputational resources
Core Design Contradiction:
Ease of operationVSUse of energy by moving object

Solution Approach 1:

The system implements partial action by providing selectable levels of NLU analysis depth. Developers can choose to implement comprehensive natural language understanding only when necessary, or use simpler matching for routine operations, allowing optimization of computational resources based on specific application needs while maintaining high user interaction quality where required.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4102502A1Service for developing dialog-driven applications
Publication Date: 2022.12.14 AMAZON TECH INC
  • EP4102502A1 patent drawingFigure 1
  • EP4102502A1 patent drawingFigure 2
  • EP4102502A1 patent drawingFigure 3

AI summary

A natural language understanding model is trained using respective natural language example inputs corresponding to a plurality of applications. A determination is made as to whether a value of a first parameter of a first application is to be obtained using a natural language interaction. Using the natural language understanding model, at least a portion of the first application is generated.