Dealing with Out of Domain Questions in Virtual Characters

Abstract

We consider the problem of designing virtual characters that support speech-based interactions in a limited domain. Previously we have shown that classification can be an effective and robust tool for selecting appropriate in-domain responses. In this paper, we consider the problem of dealing with out-of-domain user questions. We introduce a taxonomy of out-of-domain response types. We consider three classification architectures for selecting the most appropriate out-of-domain responses. We evaluate these architectures and show that they significantly improve the quality of the response selection making the user's interaction with the virtual character more natural and engaging.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2006
Accession Number
ADA459219

Entities

People

  • Anton Leuski
  • David R Traum
  • Ronakkumar Patel

Organizations

  • University of Southern California

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Abstracts
  • Automated Speech Recognition
  • Behavior And Behavior Mechanisms
  • Classification
  • Data Sets
  • Education
  • Information Retrieval
  • Language
  • Machine Learning
  • Personality
  • Probability Distributions
  • Supervised Machine Learning
  • Taxonomy
  • Test Sets
  • Training
  • United States
  • United States Government

Fields of Study

  • Computer science

Readers

  • Database Systems and Applications
  • Speech Processing/Speech Recognition.
  • Theoretical Analysis.