Using Machine Learning to Train a Wearable Device for Measuring Students' Cognitive Load during Problem-Solving Activities Based on Electrodermal Activity, Body Temperature, and Heart Rate: Development of a Cognitive Load Tracker for Both Personal and Classroom Use.

William L Romine,Tanvi Banerjee,Josephine Graft,Mahdieh Zabihimayvan,Fan Yang,Dipesh Kadariya,Reza Sadeghi,Noah L Schroeder

doi:10.3390/s20174833

Abstract

Automated tracking of physical fitness has sparked a health revolution by allowing individuals to track their own physical activity and health in real time. This concept is beginning to be applied to tracking of cognitive load. It is well known that activity in the brain can be measured through changes in the body’s physiology, but current real-time measures tend to be unimodal and invasive. We therefore propose the concept of a wearable educational fitness (EduFit) tracker. We use machine learning with physiological data to understand how to develop a wearable device that tracks cognitive load accurately in real time. In an initial study, we found that body temperature, skin conductance, and heart rate were able to distinguish between (i) a problem solving activity (high cognitive load), (ii) a leisure activity (moderate cognitive load), and (iii) daydreaming (low cognitive load) with high accuracy in the test dataset. In a second study, we found that these physiological features can be used to predict accurately user-reported mental focus in the test dataset, even when relatively small numbers of training data were used. We explain how these findings inform the development and implementation of a wearable device for temporal tracking and logging a user’s learning activities and cognitive load.

Highlights

Monitoring engagement with learning materials in real time can be challenging
We explore the efficacy of electrodermal activity, skin temperature, and heart rate for classifying three activities, which are associated with varying levels of cognitive load
We were surprised that the support vector machine (SVM) and neural networks performed relatively poorly given their capacity to model complex relationships between input features and the output classes, with classification performance paralleling that obtained from the baseline models

Summary

Introduction

Monitoring engagement with learning materials in real time can be challenging. Many of us can likely recall instances where we had to re-read a paragraph because we were thinking about something else and had no recollection of what the paragraph said. We envision a system that helps learners identify when they are not putting appropriate mental effort into learning, and believe that real-time classification of one’s cognitive load could improve study habits. Sensors 2020, 20, 4833 implementation of the Educational Fitness (EduFit) system, which utilizes physiological data from a wearable sensor and machine learning to differentiate between different levels of cognitive load for the user. While the literature shows an abundance of research focused on applying the concept of cognitive load, mental effort, or mental workload to different domains, the measurement of cognitive load has remained rife with limitations throughout the educational literature [1,2]. The use of basic physiological measures in learning contexts has drawn renewed attention by researchers, perhaps in part due to the growing availability of minimally invasive wearable sensors. A decrease in heart rate between the beginning and end of a 60-min class may suggest that students lose alertness and neurocognitive excitability [5]

Methods

Results

Conclusion