Teaching Senti-Lexicon for Automated Sentiment Polarity Definition in Teaching Evaluation

This research significantly achieved the construction of a teaching evaluation sentiment lexicon and an automated sentiment orientation polarity definition in teaching evaluation. The Teaching Senti-lexicon will compute the weights of terms and phrases obtained from student opinions, which are stored in teaching evaluation suggestions in the form of open-ended questions. This Teaching Senti-lexicon consists of three main attributes, including: teaching corpus, category and sentiment weight score. The sentiment orientation polarity was computed with its mean function being sentiment class definitions. A number of 175 instances were randomised using teaching feedback responses which were posted by students studying at Loei Raja hat University. The contributions of this paper propose an effective teaching sentiment analysis method, especially for teaching evaluation. In this paper, the experimented model employed SVM, ID3 and Naïve Bayes algorithms, which were implemented in order to analyse sentiment classifications with a 97% highest accuracy of SVM. This model is also applied to improve upon their teaching as well.