Educational Big Data Analytics Using Sentiment Analysis for Student Requirement Analysis on Courses

Abstract

The online learning become a choice of most educational institution which creates enormous data on learning platform. This study introduces a novel framework that leverages Big Data analytics, with a focus on sentiment analysis, to decipher student requirements and preferences regarding course offerings and content. The objective is to harness the vast amounts of unstructured feedback generated by students in the form of reviews, forum posts, and surveys to inform and enhance educational strategies. We propose a sentiment analysis model multi attention fusion with CNN-BiLSTM model, that is adept at processing natural language and identifying the polarity of sentiments expressed by students. By analyzing this sentiment data, our system can capture the nuanced preferences and needs of students. The model is trained and validated on a diverse dataset encompassing various educational domains and student demographics, ensuring robustness and generalizability of the results. The outcomes indicate that sentiment analysis is an effective tool for uncovering hidden patterns and trends in student feedback. Our findings reveal correlations between student satisfaction and specific course features, such as module content, teaching methodologies, and resource availability. Additionally, the results evaluate precision, recall, accuracy and F1-score.