Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Linguistic adaptations during spoken and multimodal error resolution.

S Oviatt1, J Bernard, G A Levow

  • 1Department of Computer Science, Oregon Graduate Institute of Science and Technology, Portland 97291, USA. oviatt@cse.ogi.edu

Language and Speech
|April 4, 2000
PubMed
Summary

Users adapt their speech and language to fix errors in recognition systems. They increase contrast, hyperarticulate speech, and suppress variability for better intelligibility.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Modeling global and focal hyperarticulation during human-computer error resolution.

The Journal of the Acoustical Society of America·1998
Same author

[Efficacy of radical treatment with halofantrine on the prevention of imported Plasmodium falciparum malaria].

Annales de la Societe belge de medecine tropicale·1992
Same author

Cancer cures. Yesterday, today, and tomorrow.

The Mount Sinai journal of medicine, New York·1992
Same author

Striking identity between HIV-1 envelope glycoprotein gp120 and its CD4 receptor.

Lancet (London, England)·1992
Same author

Marketing CT and MR imaging services in a large urban teaching hospital.

Radiology·1992
Same author

Postnatal changes in AMPA receptor regulation by phospholipase A2 treatment of synaptic membranes: temporally differential effects on agonist and antagonist binding.

Brain research. Developmental brain research·1992

Area of Science:

  • Human-Computer Interaction
  • Linguistics
  • Speech Recognition

Background:

  • Error handling in recognition systems is a significant challenge.
  • System failures lead to user frustration and limit commercial viability.
  • Understanding user adaptation during error resolution is crucial for system improvement.

Purpose of the Study:

  • To analyze linguistic adaptations during spoken and multimodal human-computer error resolution.
  • To identify the types and magnitude of user adaptations in response to recognition errors.
  • To investigate how users modify their input to resolve persistent errors.

Main Methods:

  • Utilized a semiautomatic simulation with error-generation capabilities.
  • Collected spoken and pen-based input samples before and after recognition errors.

Related Experiment Videos

  • Analyzed user input across different error correction depths (repetitions).

    • Main Results:

    • Users adapt by increasing linguistic contrast (alternating input modes/content).
    • Verbatim speech corrections involve hyperarticulation (lengthened segments, pauses, falling contours).
    • Hyperarticulation is accompanied by suppressed variability in amplitude and fundamental frequency.

    Conclusions:

    • User adaptations enhance linguistic intelligibility during error correction.
    • Findings support and generalize the Computer-elicited Hyperarticulate Adaptation Model (CHAM).
    • Results offer insights for developing improved error handling in future spoken and multimodal systems.