Natural language processing approaches for student evaluation analysis

Student evaluation has been critical in all educational institutions and the popularity of student feedback has increased, especially during the COVID-19 pandemic when most colleges and universities switched from traditional face-to-face instruction to an online platform. Student feedback is valued information used to improve and develop learning materials for future generations. However, interpreting student evaluation is a complicated task due to the variety of comment formats and the large volume of data, which often hides useful and valuable information. Therefore, the application of natural language processing (NLP) in analyzing student surveys has gained popularity among researchers, indicated by the growing number of studies related to the use of this technology in the education domain. One of the benefits of using NLP is the ability to process a large amount of text data in a short amount of time for effective results. This paper presents a review of some NLP technique applications on student evaluation analysis, focusing on topic modeling and sentiment analysis. It also includes implementation strategies for topic modeling models and sentiment analysis models for processing student feedback. The methods of the research are literature review and technical experiments. This research will reduce the time taken to read student feedback from the University of Tasmania (UTAS) by effectively discovering topics of a large dataset as well as assigning a sentiment score for the feedbacks.