The%20State%20of%20the%20Art%20in%20VoiceXML

1
The State of the Art in VoiceXML

• Chetan Sharma, MS Graduate Student
• School of CSIS, Pace University

2
Introduction

• VoiceXML (VXML) is a Web-based markup language
  for human-computer voice dialogs, a derivative of
  XML
• VXML was introduced with collaboration between
  IBM, Motorola, Lucent and ATT
• VXML 1.0 is a specification of VXML Forum
• VXML leverages the Internet for voice application
  development and delivery
• The VoiceXML language provides a clean and simple
  means for
• Playing audio
• Recognizing speech and touch-tone (DTMF) input
• Controlling the flow of a phone call

3
VXML Interpreter
Web Surfing
Banking
E-mail

• The heart of this system is a VXML interpreter
• Dialog is initiated when the Telephone Resource
  Manager receives an incoming call

Other Services
Paging Services
Web Server
VXML Interpreter
Other Devices
Telephone Network
4
Components of a VXML System

• Any web site can be a VXML content server
• Requires no special hardware or software
• VXML scripts are very similar to HTML documents
• Examples
• ltPROMPTgt tag indicates that the system should
  playback a piece of recorded audio
• ltFIELDgt tag is used to indicate an INPUT
  fieldVXML
• VXML Gateway controls the interpretation of the
  script and its interaction with the user
• Scripts may be embedded server-side or client-side

5
Tools for developing VXML Applications

• IBMs WebSphere Voice Server SDK
• TellMe Studio by TellMe Networks, Inc.
• Motorolas Mobile ADK
• Nuance Communications V-Builder
• BeVocal Café by BeVocal Inc.

6
IBMs WebSphere Voice Server SDK

• Leverages the multimedia capabilities
• Provides a desktop based simulation environment
  for testing voice applications
• Components
• IBMs ViaVoice Speech Recognition Engine
• Text-to-Speech Engine
• VXML Browser
• DTMF Simulator

7
TellMe Studio by TellMe Networks

• TellMe Studio consists of
• MyExtensions Application development and
  publication MyStudio Allows developers to access
• Syntax Checker
• Record by phone option
• Scratchpad or external URL
• DTMF generator

A Typical Snapshot of Flow of TellMe Platform
8
Motorolas Mobile ADK

• Components
• IDE that supports validation of VXML Application
• Desktop based VXML Simulator
• Microsofts agent-based application

9
Nuance Communications V-Builder

• Consists of
• Visual IDE to develop VXML based applications
• Fundamental elements text areas, text fields,
  buttons and menus that form the part of the IDE
• Also incorporates Nuance SpeechObjects for common
  dialogs, entry, database and web queries

10
BeVocal Café by BeVocal, Inc.

• Consists of
• File Management (for uploading grammar and audio
  files)
• VXML Checker (to validate VXML content)
• Trace Tool (for tracing and debugging apps)
• Log Browser (for viewing the call trace log)
• Port Estimator

11
Goals and Advantages of VXML

• Goals
• To bring the full power of web development and
  content delivery to voice response applications
• To enable integration of voice services with data
  services
• Advantages
• Minimizes client/server interactions
• Shields application authors from low-level, and
  platform-specific details
• Separates user interaction code (in VXML) from
  service logic (CGI scripts)
• Promotes portability across implementation
  platforms
• Safely handles shared network-based applications

12
Thank You