A novel framework for MOOC recommendation using sentiment analysis

Abstract

Massive open online courses (MOOC) are the largest initiative in eLearning, with the support of universities across the world. To increase course satisfaction in MOOCs, learners’ must relate to the courses that best suit their needs and interests. The goal of recommendation systems is to suggest items to users based on their preferences and past behaviour. A course recommender system makes recommendations based on the similarity of courses and past interactions with the MOOC platform. With a huge volume of online courses on multiple learning platforms, it has been difficult for learners to identify the course of their interest. To address these challenges, a novel framework for hybrid MOOC course recommendations is proposed to recommend courses from multiple learning platforms. It uses web scraping techniques to collect course data from various MOOC providers, such as Coursera, Udemy, and edX platforms. With the real time dataset, a deep learning chatbot captures the personalized learning requirements of learners and recommends using a user-user collaborative approach with the valence aware dictionary and sentiment reasoner (VADER) for sentiment analysis. It enhances the accuracy of recommendations with an root-mean-square error (RMSE) value of 0.541.