Students’ Perceptions of E-Learning in Malaysian Universities: Sentiment Analysis Based Machine Learning Approach

• DOI: 10.28945/5024
• Publication Date: 2022-10-08T13:13:28Z
• Authors: Reem Baragash; Hanan Aldowah; Irfan N Umar
• Keywords: Sentiment Analysis; Confusion matrix; Confusion

Aim/Purpose: To gain insight into the opinions and reviews of Malaysian university students regarding e-learning systems, thereby improving quality services these systems resolving any problems, concerns, issues that may exist within institution. Background: This exploratory study examines students’ perceptions in Malaysia based on Sentiment Analysis (SA) to a clear their feelings about related universities determine whether are positive or negative. Methodology: The data was collected from Twitter; Full Archive Search API Premium v1.1 tire chosen access tweets November 1, 2019, December 30, 2020. R programming language library package “rtweet” applied search query tweets. classify opinions, sentiment analysis-based Machine Learning (ML) with Support Vector (SVM) utilized. Rapid Miner, statistical mining tool, used accuracy ML algorithm. After preparing data, RapidMiner pre-process final 1201 sentiment, National Research Council (NRC) word-emotion lexicon detect presence eight emotions confusion matrix is classifier’s performance. Contribution: research provided evidence for effective use analysis as an indicator contribute development educational specifically, universities. Findings: Based findings, majority have opinion Precisely, results showed 65% sentiments were classified 35% Moreover, among emotions, expressed higher level trust, anticipation, joy. Recommendations Practitioners: findings could help teachers’ strengths weaknesses graphically negative feedback. These would also decision-makers educationalists be more aware (sentiments) concerns. Thus, using social media should encouraged valuable source information assist decision-making, development, performance evaluation. Recommendation Researchers: encourage other researchers apply SA approach Twitter discover users’ certain learning teaching processes. Impact Society: Our confirmed provide supportive procedures appropriate decision-making future strategies. Future Research: work can experiment classification models different algorithms well feature extraction methods compare find best improve