Text Normalization Authoring Tool for Speech Recognition

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The development and implementation of speech recognition systems require high-cost, resource-intensive processes for text normalization and grammar library creation, often necessitating advanced technical and linguistic expertise, leading to inefficiencies and inconsistencies.

Innovation Solution

A runtime framework and authoring tool that enables users to define and validate text normalization maps and grammar libraries without requiring advanced programming or linguistic skills, facilitating consistency and reducing duplication of effort through a user-friendly interface and shared runtime environments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a rule-based text normalization system is implemented using conventional methods, then text conversion from display form to spoken form can be achieved, but the process requires high-level linguistic and technical expertise, leading to high development costs and resource consumption

Engineering Contradiction:

Improvetext normalization accuracyVSAvoidauthoring process complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary tool that acts as a mediator between the linguist and the complex rule-based normalization system. This tool provides a simplified interface that automatically generates normalization rules from linguistic input, eliminating the need for linguists to directly program complex rules while maintaining high normalization accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables self-service by allowing linguists to author normalization rules using a simplified interface that automatically handles the complex technical aspects. The tool translates linguistic knowledge into executable normalization rules without requiring the linguist to have programming expertise, making the process self-sufficient and eliminating the need for specialized technical personnel.

Inventive Principle:

Principle #25Self-service

2Productivity

If text normalization rules and grammar libraries are authored separately, then each can be developed independently, but this leads to duplication of effort and inconsistencies between the two components

Engineering Contradiction:

Improvedevelopment efficiencyVSAvoidconsistency between normalization maps and grammars

Core Design Contradiction:

ProductivityVSStability of the object's composition

Solution Approach 1:

The patent merges the authoring processes for text normalization rules and grammar libraries into a single integrated tool. This allows both components to be developed simultaneously with automatic cross-referencing, eliminating duplication of effort and ensuring consistency between the normalization maps and grammar libraries through shared data structures and validation rules.

Inventive Principle:

Principle #5Merging (Combining)

3Ease of manufacture

If conventional text normalization tools are used, then text can be converted to spoken form, but the process requires advanced programming skills and technical interpretation, increasing the barrier to entry and development time

Engineering Contradiction:

Improveease of creating normalization rulesVSAvoidtechnical expertise required

Core Design Contradiction:

Ease of manufactureVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary authoring tool that shields users from technical complexity by providing a high-level interface. This tool automatically handles the translation from linguistic rules to technical implementations, eliminating the need for users to understand complex programming concepts while maintaining the ability to create sophisticated normalization rules.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system uses templates and reusable rule patterns that can be copied and adapted for different normalization scenarios. This allows users to leverage pre-built rule structures without understanding the underlying complexity, reducing the technical barrier to entry while maintaining flexibility and precision in text normalization.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS8990088B2Tool and framework for creating consistent normalization maps and grammars

Publication Date: 2015.03.24 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A runtime framework and authoring tool are provided for enabling linguistic experts to author text normalization maps and grammar libraries without requiring high level of technical or programming skills. Authors define or select terminals, map the terminals, and define rules for the mapping. The tool enables an author to validate their work, by executing the map in the same way the recognition engine does, causing consistency in results from authoring to user operations. The runtime is used by the speech engines and by the tools to provide consistent normalization for supported scenarios.