<![CDATA[The growth of electric vehicle (EV) adoption has generated extensive public discussions on digital platforms, reflecting diverse perceptions of environmental benefits, economic feasibility, technological readiness, and government policies. This study aims to develop a Transformer-based Natural Language Processing (NLP) framework for classifying public sentiment toward EV adoption using textual data from social media and digital platforms. A quantitative experimental approach was applied through data collection, preprocessing, sentiment and emotion labeling, Transformer-based modeling, and performance evaluation. The dataset consisted of 5,000 public opinion texts, of which 3,785 records were retained after data cleaning and selection. Sentiment classification included three categories: positive, negative, and neutral, while emotion classification consisted of happy, trust, angry, fear, disappointed, and surprise. Model performance was evaluated using accuracy, precision, recall, and F1-score. The results showed that positive sentiment was dominant, accounting for 43.46% of the analyzed opinions, followed by negative sentiment at 31.97% and neutral sentiment at 24.57%. Positive opinions were mainly related to environmental benefits and energy efficiency, whereas negative opinions reflected concerns about vehicle prices, charging infrastructure, charging time, and battery replacement costs. These findings indicate that Transformer-based NLP can capture contextual semantic information from large-scale public opinion data and support reliable sentiment classification. The proposed framework provides practical value for policymakers, researchers, and industry stakeholders in developing data-driven strategies to promote EV adoption and sustainable transportation.]]>
Copyrights © 2026
