Analisis Sentimen Masyarakat terhadap Kebijakan Gubernur Jawa Barat di Sosial Media Menggunakan Algoritma Long Short-Term Memory

This study examines public sentiment toward the policies of the Governor of West Java as expressed on social media platforms, including Twitter, Instagram, and TikTok, using the Long Short-Term Memory (LSTM) algorithm. The main challenge lies in the dynamic and dispersed nature of public opinion across multiple platforms, making conventional survey methods less effective in capturing real-time and comprehensive public perceptions.The dataset was collected through web scraping and consists of Indonesian-language comments. The data were processed using the Knowledge Discovery in Databases (KDD) framework, which includes data selection, preprocessing, transformation, labeling, and word embedding. The LSTM model was trained to classify sentiments into three categories: positive, negative, and neutral. The model's performance was evaluated using accuracy, precision, recall, and F1-score metrics.The results demonstrate that the proposed model is capable of classifying public sentiment effectively and provides valuable insights into the distribution of public opinion regarding various gubernatorial policies. These findings are expected to contribute to the development of an AI-based public opinion monitoring system that supports more responsive and evidence-based decision-making by regional governments.