Automatic Detection of a Student's Affective States for Intelligent Teaching Systems.

Mark H Myers

doi:10.3390/brainsci11030331

Abstract

AutoTutor is an automated computer tutor that simulates human tutors and holds conversations with students in natural language. Using data collected from AutoTutor, the following determinations were sought: Can we automatically classify affect states from intelligent teaching systems to aid in the detection of a learner’s emotional state? Using frequency patterns of AutoTutor feedback and assigned user emotion in a series of pairs, can the next pair of feedback/emotion series be predicted? Through a priori data mining approaches, we found dominant frequent item sets that predict the next set of responses. Thirty-four participants provided 200 turns between the student and the AutoTutor. Two series of attributes and emotions were concatenated into one row to create a record of previous and next set of emotions. Feature extraction techniques, such as multilayer-perceptron and naive Bayes, were performed on the dataset to perform classification for affective state labeling. The emotions ‘Flow’ and ‘Frustration’ had the highest classification of all the other emotions when measured against other emotions and their respective attributes. The most common frequent item sets were ‘Flow’ and ‘Confusion’.

Highlights

Intelligent tutoring systems have been the ultimate goal in remote learning paradigms for several years
Automatic detection of learning states is part of an intelligent system, where the student’s emotion or affective state is tied to a learning session and is updated to the subsequent automated learning response of the student, such as confusion, high interest, etc. [1,2]
This study focuses deeper into the physiological aspect of learning where cognitive load, heart rate, etc. are classified into groups, i.e., Effortless behavior—Observed: high mental workload, high load on memory and low heart rate (HR) (C1), Effortful behavior—Observed: high electrodermal activity (EDA), high emotion, high HR, low load on memory and low mental workload (C3), etc. [14]

Summary

Introduction

Intelligent tutoring systems have been the ultimate goal in remote learning paradigms for several years. Automatic detection of learning states is part of an intelligent system, where the student’s emotion or affective state is tied to a learning session and is updated to the subsequent automated learning response of the student, such as confusion, high interest, etc. The system may respond to the student in answering a question to the student, correcting student’s errors, or providing hints or pumps (tell me more) to enable the student to answer the question correctly. Throughout this session, affective states are captured, such as through the method known as emote aloud [3]. The experimental result indicated that the recognition rate of the experiment, which used the mouse movement data with the facial labeled data, is higher than just the facial labeled data (94.60% vs. 91.51%)

Methods

Results

Discussion

Conclusion