Manuscript received on April 18, 2020. | Revised Manuscript received on July 22, 2020. | Manuscript published on April 30, 2020. | PP: 665-668 | Volume-9 Issue-4, April 2020. | Retrieval Number: D7573049420/2020©BEIESP | DOI: 10.35940/ijeat.D7573.049420
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Student’s feedback is a very important and crucial tool for any teacher to know his/her performance and to plan the potential improvements. In the present study the student’s opinion comprises of the several characteristics about their teacher’s teaching performance are collected through closed ended questions on Likert scale as well as through few open ended questions in terms of brief statements. The open ended questions being difficult to infer; mostly are ignored. For such cases the sentiment analysis is a good tool to bring them in main analysis stream for the inference of data. The satisfaction level pertaining to the teacher’s overall performance also is enquired during the survey as a datum reference. Multinomial logistic regression is fitted and satisfaction levels are estimated using it. These two values are compared subsequently with the third resultant value that is obtained by using the sentiment analysis to evaluate, whether the same feelings about a teacher are being reflected from students feedback collected in the form of statements or not. Sentiment analysis is a technique used to measure the sentiments in numerical values that are associated with the text or the statements under consideration. The sentiment scores obtained using sentiment analysis is further processed to estimate the satisfaction levels and classification rates. Further artificial neural network is used to find out the important characteristic which characterizes the satisfaction levels.
Keywords: Feedback, sentiment analysis, multinomial logistic regression, artificial neural network, classification, word cloud, classification rates