Sentiment Analysis on Smart City Mobile Platform Based on Lexicon

This study investigates the analysis of social media data from platforms such as Facebook, Twitter, Instagram, and application reviews, with a focus on Smart City mobile platforms. While the availability of public service-related data presents a promising opportunity for analysis, the complexities and nuances of language pose significant challenges. To overcome this, the study employs lexicon-based analysis and machine learning-based classifiers to examine sentiment on Smart City platforms using data from reviews on the Tangerang Live application. The results demonstrate that lexicon-based analysis accurately describes the weighting of each word, enabling a clear portrayal of sentiment distribution. Furthermore, sentiment results can serve as a foundation for labeling, as evidenced by the high accuracy of the random forest, k-nearest neighbors, and naive Bayes classifiers, achieving 84%, 72%, and 64% accuracy, respectively. Overall, this study offers valuable insights into sentiment analysis on Smart City platforms, which can inform future research in this area.