Automated Prompt Collection for Dynamic Voice Applications

Solution Overview

Problem

Developers face challenges in replacing Text-to-Speech (TTS) generated audio with professionally recorded audio in dynamically generated voice-enabled web applications, as existing methods require manual identification and isolation of prompts, which is cumbersome and inefficient.

Innovation Solution

A method and system using a prompt collection tool to automate the process of identifying and replacing TTS generated audio, involving a voice-enabled application server, VoiceXML Simulator, and processor to capture and record audio prompts, delegate text-to-speech services, and push Speech Synthesis Markup Language text to a prompt collecting tool.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If developers manually identify and isolate prompts by going through each line of code, then they can replace TTS with professional recordings, but the process becomes very difficult and time-consuming especially when prompts are dynamically generated

Engineering Contradiction:

Improveprompt identification accuracyVSAvoidprompt collection time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The system enables self-service by allowing the voice application to automatically identify and collect its own prompts through instrumentation. The prompt collection tool intercepts TTS requests directly from the application code, eliminating the need for developers to manually trace through dynamic code paths and extract prompts themselves.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

A prompt collection tool is introduced as an intermediary component between the voice application and the TTS engine. This mediator automatically captures all TTS requests regardless of how dynamically they are generated, replacing the manual code analysis approach and enabling systematic collection of all prompts across all application paths.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If developers deploy the application with audio files and listen to every path to manually identify TTS generated audio, then they can replace TTS with professional recordings, but the process becomes burdensome and inefficient

Engineering Contradiction:

Improveprompt identification accuracyVSAvoidprompt collection ease

Core Design Contradiction:

Manufacturing precisionVSEase of operation

Solution Approach 1:

The system enables self-service by allowing the voice application to automatically identify and collect its own prompts through instrumentation. The prompt collection tool intercepts TTS requests directly from the application code, eliminating the need for developers to manually trace through dynamic code paths and extract prompts themselves.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The manual mechanical process of listening to audio paths and identifying prompts is replaced with an automated software-based system. The prompt collection tool uses code instrumentation to automatically detect and capture TTS requests, substituting the manual auditory identification process with automated programmatic detection.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If professionally recorded audio is used instead of Text to Speech generated audio, then customer satisfaction and corporate brand image are improved, but the complexity of capturing all dynamically generated prompts increases significantly

Engineering Contradiction:

Improvecustomer satisfactionVSAvoidprompt capture system complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

A prompt collection tool is introduced as an intermediary component between the voice application and the TTS engine. This mediator automatically captures all TTS requests regardless of how dynamically they are generated, replacing the manual code analysis approach and enabling systematic collection of all prompts across all application paths.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The prompt collection functionality is extracted as a separate, dedicated tool that can be independently deployed and configured. This extraction allows the complex task of prompt capture to be handled by a specialized component rather than being embedded throughout the entire voice application codebase, reducing overall system complexity.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS8126716B2Method and system for collecting audio prompts in a dynamically generated voice application

Publication Date: 2012.02.28 CERENCE OPERATING CO

AI summary

A prompt collecting tool (190) for an interactive voice response system (100) includes a voice enabled application server (150), a voice simulator coupled to the voice enabled application server, and a processor coupled to the voice simulator. The processor can be programmed to execute (202) a voice application having a plurality of audio prompts, play (206) audio if a pre-stored audio is available for a particular prompt, capture (208) text when no pre-stored audio is available and forward (210) the captured text to the prompt collecting tool. The voice simulator can include a VoiceXML browser (160), a text to speech text service (170), and a text based recognition service (180) for example.