A sentiment analysis model to analyze students reviews of teacher performance using support vector machines

Abstract

Teacher evaluation is considered an important process in higher education institutions to know about teacher performance and implement constructive strategies in order to benefit students in their education. The opinion of students has become one of the main factors to consider when evaluating teachers. In this study we present a Model called SocialMining using a corpus of real comments in Spanish about teacher performance assessment. We applied Support Vector Machines algorithm with three kernels: linear, radial and polynomial, to predict a classification of comments in positive, negative or neutral. We calculated sensibility, specificity and predictive values as evaluation measures. The results of this work may help other experiments to improve the classification process of comments and suggest teacher improvement courses for teachers.