Student Evaluations in Teaching – Emotion Classification Using Neural Networks

Abstract

Student evaluation of teaching effectiveness plays an important role in Higher Education. Evaluations serve as Formative (identify areas of improvement in the process) and Summative (assess the end goal) measurements of teaching. Educational institutions collect these evaluations in both qualitative and quantitative forms. Qualitative evaluations serve as a bridge for students to express their feelings about the teaching methodology used, instructor efficiency, classroom environment, learning resources and others. Identifying student emotions help instructors to have good intellectual insight about the actual impact of teaching. Teaching models include traditional models, modern flipped class-room models, and active learning approaches. Light-weight team is an active learning approach, in which team members have little direct impact on each other’s final grades, with significant long-term socialization. We propose and extend previous method for assessing the effectiveness of the Light-weight team teaching model, through automatic detection of emotions in student feedback in computer science course by using Neural Network model. Neural Networks have been widely used and shown high performance in variety of tasks including but not limited to Text Classification and Image Classification. It is highly deemed to work great with huge volume of data. In this study we discuss how sequential model can be used with smaller data sets and it performs well, compared to the baseline models such as Support Vector Machines and Naive Bayes.



Author Information
Jaishree Ranganathan, University of North Carolina at Charlotte, United States
Angelina Tzacheva, University of North Carolina at Charlotte, United States

Paper Information
Conference: ECE2020
Stream: Higher education

This paper is part of the ECE2020 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window

Video Presentation

Posted by amp21