A Method for Estimating Whether a User is in Smooth Communication with an Interactive Agent in Human-Agent Interaction

Takanori Komatsu, Shoichiro Ohtsuka, Kazuhiro Ueda, Takashi Komeda, Natsuki Oka

Research output: Chapter in Book/Report/Conference proceedingChapter

1 Citation (Scopus)

Abstract

The purpose of this paper is to propose a criterion for evaluating whether users are in smooth communication with an interactive agent or not by means of cognitive, behavioral and physiological indexes of these users: To do so, we conducted an experiment to measure the above three indexes of subjects who experienced smooth interactions (Group 1) or disrupted interactions (Group 2) with an experimenter in a simple cooperating game. As a result, there were no significant differences in cognitive index (subjective report: NASA-TLX score) between two subject groups, while there were in behavioral (performance of subjects: game score) and physiological indexes (heart beat variability: RSA and MWSA). These results showed that the monitoring the behavioral and physiological indexes are available for evaluating their communication and leads to acquire a desired criterion for doing so.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsMircea Gh. Negoita, Robert J. Howlett, Lakhmi C. Jain
PublisherSpringer Verlag
Pages371-377
Number of pages7
ISBN (Print)9783540301325
DOIs
Publication statusPublished - 2004 Jan 1

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3213
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'A Method for Estimating Whether a User is in Smooth Communication with an Interactive Agent in Human-Agent Interaction'. Together they form a unique fingerprint.

Cite this