Improvement of the Teaching–Learning Process Using Feature-Driven Opinion Mining of Stakeholders Comments

Higher education institutions have grown interested in enhancing the standard of the educational process by employing comments to track instructor’s instructions, student achievements, and curriculum assessment. Students, instructors, and various other participants must participate fully in the teaching–learning process of education based on outcomes in order to identify and assess many facets of education. Employing various computerized communication mediums, a significant number of perspectives on matters relevant to learning are shared on a daily basis in a modern, digitally connected society. A possible starting point for assessing the teaching–learning process might be found in the comments that learners and instructors have posted. Whenever it is performed manually, managing this much material is once again an exhausting and time-consuming process. Additionally, it is quite challenging to draw out thoughts regarding various features of recorded, unstructured language. Through social networking sites, an enormous amount of context and justification are presented every day. One of the most popular methods for extracting feelings from unstructured text is opinion analysis. Sentimental analysis of social media websites was done to reduce the use of the old methods for gathering suggestions and comments. The majority of the investigation carried out has only been used to evaluate student’s reviews and categorize the sentiment as favorable or non-favorable, utilizing linguistic or supervised approaches and considering the whole document as one. The study has discovered that opinion extraction is a significantly underutilized instrument in academic affairs for determining insights on various topics. The results using precision and recall of the proposed research show that a feature-driven opinion extraction approach, in which sentiments are extracted for aspects being discussed, enhanced the process of teaching and learning.