Handling Out-of-Grammar Commands in Mobile Speech Interaction Using Backoff Filler Models

Abstract

In command and control (C and C) speech interaction, users interact by speaking commands or asking questions typically specified in a context-free grammar (CFG). Unfortunately, users often produce out-of-grammar (OOG) commands, which can result in misunderstanding or nonunderstanding. We explore a simple approach to handling OOG commands that involves generating a back off grammar from any CFG using filler models, and utilizing that grammar for recognition whenever the CFG fails. Working within the memory footprint requirements of a mobile C and C product, applying the approach yielded a 35% relative reduction in semantic error rate for OOG commands. It also improved partial recognitions for enabling clarification dialogue.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2007
Accession Number
AD1170991

Entities

People

  • David M. Chickering
  • Sudeep Gandhe
  • Tim Paek
  • Yun C. Ju

Organizations

  • University of Southern California

Tags

Communities of Interest

  • C4I

DTIC Thesaurus Topics

  • Accuracy
  • Automated Speech Recognition
  • Command And Control
  • Computer Vision
  • Context Free Grammars
  • Data Sets
  • Errors
  • Grammars
  • Language
  • Mobile Devices
  • Mobile Phones
  • Natural Language Processing
  • Natural Languages
  • Recognition
  • Simulations
  • Spreadsheet Software
  • Syntax
  • Test And Evaluation

Fields of Study

  • Computer science

Readers

  • Artificial Intelligence
  • Computational Linguistics
  • Parallel and Distributed Computing.

Technology Areas

  • Fully Networked C3
  • Fully Networked C3 - Command and Control