Larson Technical Services: Voice Technology Consulting

Developing Speech Applications

(0.5 day course)

Learning Objective

Learn the steps and processes for developing speech applications

Target Audience

Managers, project leaders, and developers of speech applications

Instructor Background

Dr. Jim Larson is the co-chair of the W3C Voice Browser Working Group, which recently adopted VoiceXML as the basis for its W3C Speech User Interface Framework. He is the author of the book, VoiceXML—Introduction to Developing Speech Applications. Dr. Larson is the manager of Advanced Human Input/Output at Intel Corporation. Jim is a certified VoiceXML 2.0 developer.

Detailed Description

Motivation for Speech Applications

  • Users access Web sites from any telephone, anywhere, anytime.
  • Speaking and listening are the natural usage modes for phones
  • When speech applications are NOT appropriate

Process for building a voice application

  • Who’s who—The playbill for the speech team
  • Investigation stage—Identify the application-focus groups, ethnography, and market analysis
  • Design stage—Specify the application-scope, functions, conceptual model, and persona
  • Choose the right technology—Evaluation criteria, comparison shopping
  • Development stage—Develop the application-user testing, iterative refinement sustaining stage; deploy and monitor the application; adjust to usage changes
  • How to allocate effort for each stage

The tasks unique to speech application development and how to plan for them

  • How to design a persona appropriate to the application and business
  • How to specify dialogs, including prompts that encourage users to speak, grammars that define how users respond to prompts, and event handlers that assist users who fail to respond to prompts appropriately
  • How to conduct Wizard of Oz tests to refine dialog designs
  • How to conduct usability tests to make a good speech application into a world-class application
  • How to monitor the application to detect changes in usage patterns and how to deal with them
  • Choosing between buying a speech server vs. using a hosting service

How to manage the customer

  • Assisting the customer to create dialog scenarios that drive the voice user interface design (and limit the scope of the user interface)
  • Specifying performance and preference criteria that drive the optimization of the user interface (and set appropriate customer expectations)
  • Helping the customer understand the true cost of last minute changes (and how to limit last minute changes)

Summary and Q&A

Comments are off for this post