Toward an Affect-Sensitive AutoTutor

  • Authors:
  • Sidney D'Mello;Rosalind W. Picard;Arthur Graesser

  • Affiliations:
  • University of Memphis;MIT Media Laboratory;University of Memphis

  • Venue:
  • IEEE Intelligent Systems
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Emotions (affective states) are inextricably bound to the learning process, as are cognition, motivation, discourse, action, and the environment. Augmenting an intelligent tutoring system with the ability to incorporate such states into its pedagogical strategies can improve learning. Two studies use observational and "emote aloud" protocols to identify learners' affective states as they interact with AutoTutor. A third study uses sensors to collect training and validation data during AutoTutor sessions through learners' conversational cues, gross body language, and facial expressions. By adapting its instructional strategies, an affect-sensitive AutoTutor could promote learning. This article is part of a special issue on intelligent educational systems.