Understanding Public Perception Made Easy: A Sentiment Analysis of Public Transportation Services

Abstrak

Public policy evaluations are often constrained by traditional methods that require significant time and resources, limiting their timeliness and impact. This research explores the use of X (formerly Twitter ) sentiment analysis to evaluate public perceptions of TransJakarta services, addressing the shortcomings of conventional survey-based approaches. The study contributes to academic literature on big data in public administration and offers policymakers a faster and more inclusive method of capturing citizen perspectives. Using X data collected from September 2023 to May 2024, sentiment analysis was conducted using VADER and TextBlob, supported by complementary techniques including word frequency analysis, word clouds, and comparative analysis. The findings reveal that public sentiment fluctuates in response to service disruptions. Notable discrepancies between the VADER and TextBlob classifications indicate the value and necessity of manual validation. In contrast to earlier studies that employed complex models less accessible to practitioners, this study presents a simplified yet robust approach to big-data-driven evaluation, making sentiment analysis more practical for policy monitoring and improvement.