A Sentiment Analysis Model for Faculty Comment Evaluation Using Ensemble Machine Learning Algorithms

Abstract

Teacher evaluation is the systematic procedure done in educational institutions to review the performance of the teachers in a classroom. It aims to provide constructive feedback for teacher's professional growth which benefits students in their education. Students' feedback in the evaluation typically include textual comments which are unstructured but are rich with adequate information and insight about teacher's mastery of the course, teaching style, course content and learning experiences of the students. In this study, sentiment analysis or opinion mining was used to analyze the students' comments. An ensemble approach integrating five individual machine algorithms namely Naive Bayes, Logistic Regression, Support Vector Machine, Decision Tree and Random Forest algorithms were applied to classify the comments based on Majority Voting Principle. The experimental result shows that the ensemble classification system outperforms these individual classifiers with 90.32% accuracy. It helps to improve machine learning results producing better predictions compared to a single model.