Text Analysis of Teaching Evaluation Based on Machine Learning

Abstract

The traditional teaching quality evaluation methods of colleges and universities have been unable to meet the informatization and modern teaching modes in terms of accuracy and implementation efficiency. Therefore, for the problem of evaluating teaching quality in colleges and universities, this paper proposes a sentiment analysis method for teaching evaluation text based on machine learning. This article establishes a teaching evaluation feature dictionary, reduces the dimensionality of attribute features through mining analysis, and extracts the features most relevant to teacher evaluation. In addition, the support vector machine algorithms with linear kernel, polynomial kernel and radial basis kernel are used to classify the sentiment of the text data in teaching evaluation to judge the sentiment tendency of evaluation. The experimental results show that the support vector machine radial basis kernel has the best effect on the classification of teaching evaluation text data, which can enable teachers to accurately obtain feedback information for evaluation, so that they can adjust their teaching work in time to assist teaching decisions and improve teaching quality.