Driving Change or Doubt? A Sentiment Analysis of Malaysian Reactions to Electric Vehicles

Electric vehicles (EVs) are gaining traction globally as a cleaner alternative to traditional fuel-powered cars. In Malaysia, while awareness is growing, public sentiment remains mixed shaped by concerns such as environmental benefits, pricing, infrastructure, and the risk of floods. This study investigates Malaysian public sentiment towards EV adoption through YouTube comments posted between 2020 and 2024, a period marked by significant EV development. The main objectives are to assess the nature of these sentiments (positive, neutral, negative) and to identify which machine learning classifier, Logistic Regression, Naive Bayes, or K-NN performs best in sentiment classification. Data was extracted using the YouTube API and preprocessed through techniques such as text cleaning, tokenization, and TF-IDF feature extraction. Sentiment polarity was determined using a lexicon-based method in RapidMiner, and classification models were evaluated using accuracy, precision, and recall. The findings revealed that neutral sentiment dominated (49.85%), followed by positive (35.58%) and negative (14.57%). Topic modelling via Latent Dirichlet Allocation (LDA) showed that issues like battery performance, charging infrastructure, and pricing were major areas of concern. Among the classifiers tested, Logistic Regression yielded the highest accuracy. This study recommends enhanced public communication on EV affordability and infrastructure readiness, especially in flood-prone areas. The insights can inform policymakers and industry players on how to address public concerns and accelerate sustainable transport adoption in alignment with SDG 11. Understanding public sentiment is not just technical, it is essential for building climate-resilient, future-ready urban mobility in Malaysia.