Faculty Advisor or Committee Member

Ivon Arroyo, Advisor

Identifier

etd-3766

Abstract

An important goal of Educational Data Mining is to provide data and visualization about students’ state of knowledge and students’ affective states. The combination of these provides an understanding of the easiness or hardness of the concepts being taught and the student’s comfortability in it. While various studies have been conducted on estimating students’ knowledge and affect, little research has been done to transform this collected (Raw) data into meaningful information that is more relatable to teachers, parents and other stakeholders, i.e. Non-Researchers. This research seeks to enhance existing Teacher Tools (An application designed within the MathSpring - An Intelligent Tutoring system) to generate a live dashboard for teachers to use in the classroom, as their students are using MathSpring. The system captures student performance and detects students’ facial expressions, which highlight students emotion and engagement, using a deep learning model that detects facial expressions. The live dashboard enables teachers to understand and juxtapose the state of knowledge and corresponding affect of students as they practice math problem solving. This should help teachers understand students’ state of mind better, and feed this information back to act and alter their instruction or interaction with each student in a personalized way. We present results of teachers' perceptions of the usefulness of the Live Dashboard, through a qualitative and quantitative survey.

Publisher

Worcester Polytechnic Institute

Degree Name

MS

Department

Computer Science

Project Type

Thesis

Date Accepted

2020-05-12

Accessibility

Unrestricted

Subjects

Facial Expression, Facial Expression Predictor, Emotion Analysis, Live Dashboard, MathSpring

Share

COinS