Training

 
Course Description

Application Tuning and Analytics Using Nuance Tools

This course provides extensive hands-on exercises to explore the strategies for identifying caller experience and recognition success in a system overall and for isolating potential problems by using two key tuning tools available with the Nuance Recognizer. We discuss best practices for application tuners, project managers, and others working on a speech team. The Nuance Application Reporting (NAR) tuning tool offers approximately 50 reports to effectively query deployment data logged by the Nuance Recognizer and Nuance Dialog Modules. These reports offer important options for viewing and analyzing application data to evaluate application performance. The Nuance Tune tool compares transcriptions of caller input to recognition results returned by the Nuance recognizer and to application grammars, and tells us if the recognizer was correct. Results from Nuance Tune are fed back in to the NAR tool for analysis of ‘out of vocabulary’ input and to help reconfigure confidence thresholds, when desirable. The course discusses application analysis with emphasis on:

  • Call analysis to identify success and failure patterns and explore root causes
  • Analysis of each dialog module’s performance including success and frequency of nomatch, noinput and confirmation prompting
  • Transaction completion analysis
  • Measurement of recognizer effectiveness and efficiency
  • End-call analysis to explore how, when and why callers hang up or transfer to the call center.
  • Transcription and Nuance_Tune

    Activities include review of individual reports within the NAR tool and exercises via Web-accessed virtual machines that allow participants to work with the NAR tool.

    This workshop is offered live, online using web conferencing software and VoIP. Participants are encouraged to have a headset with microphone to enable active discussions. Participants will also use virtual machines accessed via URL to perform class exercises.

Objectives
At the end of this course, participants should be able to:
  • Explain the goal of speech application analysis
    • View and interpret Nuance Application Reporting reports to
    • Perform Call Analysis
    • Perform Call End Analysis
    • Perform Task Analysis
    • Perform Dialog State Analysis
  • Use Nuance Tune to
    • Perform Recognition Analysis
Audience
  • Application Tuners
  • Business Analysts
  • Project Managers
  • Voice User Interface Designers
  • Grammar Developers
Outline
  • Introduction
  • Module 1 – What is Analysis?
    Become familiar with the basics of speech application analysis
  • Module 2 – Overview of Primary NAR Reports
    View the key NAR reports
  • Module 3 – Performing Call Analysis
    Find successful and failed calls
  • Module 4 – Performing Call End Analysis
    Use several reports to determine how calls end
  • Module 5 – Performing Task Analysis
    Investigate what callers try to do
  • Module 6 – Performing Dialog State Analysis
    Illuminate findings from other analysis
  • Module 7 – Overview of Primary Nuance Tune Reports
    View the key Nuance Tune reports
  • Module 8 – Performing Recognition Analysis
    Look at grammar and recognizer performance
  • Module 9 – Course Wrap-up
Prerequisites
Each participant should have taken at least one of the following NSU classes (or be able to demonstrate comparable experience):
    Nuance Speech University QuickStart Package
or
  • Introduction to Speech Recognition
  • Application Tuning Overview
Price
CoursePrice: 1998 USD For classes outside of North America, prices should be converted to Euros or British Pounds, when appropriate.
Please see below for the price of each class scheduled, or see our pricing discussion under FAQs
Duration
4 Days
Package Information
Development: Call Steering   
Available Dates

No Classes with Open status

Contact us for scheduling this class.