<![CDATA[The rapid growth of social media has driven a surge in public opinion across digital platforms, including YouTube. The transportation electrification policy, particularly the adoption of electric bicycles in Indonesia, has elicited a wide range of public responses. This study aims to build a web-based sentiment analysis system that classifies YouTube comments related to this policy using the BERT (Bidirectional Encoder Representations from Transformers) algorithm. The system was developed using the waterfall model and covers text preprocessing (text cleaning, case folding, normalization, and tokenization), sentiment classification into positive, neutral, and negative categories, and interactive visualization of the results. The dataset consists of 1,000 comments from 8 YouTube videos collected through web scraping period of 2022, 2023 and 2024. The BERT model achieved an accuracy of 91.50%, with a macro precision of 88.34%, a macro recall of 90.80%, and a macro F1-Score of 89.55%. The classified sentiment distribution comprises 201 positive comments (20.1%), 645 neutral comments (64.5%), and 154 negative comments (15.4%). The web-based system integrates the entire workflow, from dataset upload and preprocessing to sentiment analysis and visualization, and can therefore support data-driven decision-making for policymakers.]]>
Copyrights © 2026
