Sentiment Analysis using Deep Learning for Students’ Feedback: A Survey

Abstract

Over the course of the past ten years, sentiment analysis has seen widespread use across a variety of fields, including education, social networks, and business. The application of sentiment analysis is developing but remains tough, in particular in the field of education, where dealing with and processing students’ thoughts is a complicated undertaking due to the nature of the language used by students and the huge volume of information. The state of the use of sentiment analysis in this field is revealed through multiple literature evaluations, each of which examines the topic from a unique vantage point and within a unique set of circumstances. However, the existing body of research does not include a review that systematically classifies the research and results of applying natural language processing (NLP), deep learning (DL), and machine learning (ML) solutions for sentiment analysis in the education domain. This is a significant gap in the field. The Internet is one of the most common mediums for sharing and exchanging thoughts, feedback, or information regarding specific subjects. Most of the time, the feedback is provided in the form of number ratings and text. Numerical ratings can be processed with ease, but there is a vast amount of unstructured textual data present on the internet. Currently, we are working on a project in which a client can propose his idea on our website and take opinions in the form of textual review. This helps him to take specific opinions from reviewers, which will be helpful in making a decision that is more appropriate. Within the scope of this paper, we analysed and compared a variety of techniques that are utilised during the processing of text data. We were able to identify issues and scope within the textual feedback analysis with the assistance of the various methodologies.