Opinion Mining on Student Feedback Data in the Education Sector


Opinion mining is for identifying a sentence’s subjective information, such as its good, negative, or neutral sentiments and it is one of the challenges of text classification. Sentiment analysis will take context information from a text and use machine learning or deep learning to calculate various algorithms and calculations to estimate the sentiment of the text. In the realm of sentiment analysis, there are many different approaches we might take to finish the job. Additionally, the platform that we wish to use to assess the sentiment varies. Opinion mining is the one of the most popular Natural Language Processing (NLP) applications to extract human thoughts from reviews. Opinion mining is used in the educational industry to hear student comments and pedagogically improve learning-teaching procedures. With enhancement in sentiment footnote techniques and Artificial Intelligence methodology, student responses can be annotated with their sentiment point of reference without greatly human involvement. The flow diagram for opinion mining of student feedback is included in the figure1.

Flow Diagram for opinion mining of Student Feedback

  1. Data Collection – Collecting feedback from the students.
  2. Data Pre-processing – Cleaning the data using tokenization, normalization, stemming and removal of irrelevant content. (pre-processing methods are not limited).
  3. Sentiment Identification – Identification with the help of NRC Lexicon. Here the NRC Lexicon means (National Research Council of Canada) Lexicon. It has list of English words and basic emotions like anger, trust, fear, joy and surprise etc.
  4. Data Visualization – Graphical representation of the feedback will be displayed after computing satisfaction and dissatisfaction from the student’s feedback.

The Role of Opinion Mining in Education Sector

Sentiment analysis has been widely used by the research community to get the feedback from the students on instructional strategies used in the education division. The opinion mining process can be modified for use in a variety of educational domain applications, including assessing student engagement pedagogy, evaluating courses and teachers, making decisions about educational legislation, and more. The above mentioned applications can be proficient by the identification of student opinion on a particular subject in a semester, and their involvement in both offline classes and online forums. The important algorithms for different applications of Educational Sectors are mentioned below:

  1. Learning and Teaching Systems Evaluation – Association rule mining and Ensemble learning,
  2. Enhance Pedagogical Concepts – Reinforcement learning and Topic Ontology
  3. Decision Making – Support Vector Machine and Artificial Neural Network
  4. Assessment Evaluation – k-nearest neighbour  and Additive regression

Challenges of Opinion Mining in Education Sector

The biggest difficulty in sentiment analysis in the educational sector is detecting opinion spam, in spite of the fact that there are many other difficulties. Opinion spam identification is crucial in the educational sector since students may use intermediary services for evaluation items and submit false evaluations to registered courses in the semester end. The scheme of detecting opinion spam has been extensively researched in the context of product comment, where people are rewarded for producing fictitious ratings.


In order to keep up with standards in today’s tough world, the education sector is undergoing enormous transformation. Positive reforms are being incorporated into every facet of education. Smooth analysis in the education sector would involve other Natural Language Processing methods even if sentiment analysis is straightforward discovery of the emotion direction of students towards the teaching or course materials. It is necessary to use learning algorithms of Artificial Intelligence to analyze the opinions of a huge number of pupils.

Senbagavalli M, Associate Professor, Department of Information Technology, Alliance College of Engineering and Design, Alliance University, Bangalore.

Saswati Debnath, Assistant Professor, Department of Computer Science & Engineering, Alliance College of Engineering and Design, Alliance University, Bangalore.