b-LSTM enabled Assessment Model for Virtual Learning Eco-system

  • Gaurav Srivastav, Shri Kant
Keywords: E-learning, VLE, Natural Language Processing, classification, LSTM, context, language modeling, BERT


In this research paper, the recent emerging trends and challenges in Virtual Learning Environment (VLE) assessment have been addressed. This paper includes a comprehensive experiment set-up for handling the challenge. Proposed model analyses the Student’s Review for a particular course in which they are enrolled. Then perform language modeling on the review using Long Short-Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers (BERT). This model attempts to understand the sentiments of the learner and predicts that a student will get an upgrade or not for the next term. If upgradation is not suitable according to the model, then subsequent tasks can be assigned to the student for improving the VLE outcomes. Based on the systematic experiment 97% accuracy has been achieved in classifying students according to their reviews and score-based assessment. The experiment has been performed over different batch sizes and outcomes are tabulated in the subsequent sections

How to Cite
Shri Kant, G. S. (2022). b-LSTM enabled Assessment Model for Virtual Learning Eco-system. Design Engineering, (1), 321-337. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/8810