Analisis Sentimen Terkait Keberadaan Tukang Parkir Liar Pada Platform X Berbasis Teknik Klasifikasi

In the digital era, social media plays a crucial role as a platform for the public to voice complaints about various issues, including parking problems. This study analyzes how the general public perceives the presence of illegal parking attendants on the social media platform X using three sentiment analysis techniques: Random Forest, Long Short-Term Memory (LSTM), and
IndoBERTweet. To construct a comprehensive dataset, this research integrates feature selection, data transformation, and preprocessing methods. The selected models are used to predict the outcomes of public complaints, categorized into “Positive,” “Negative,” “Neutral,” and “Others” sentiments. The performance of
the models is evaluated using classification metrics such as accuracy, precision, recall, F1-score, and a confusion matrix. The results of this study achieve optimal accuracy in classifying public sentiment toward illegal parking attendants. In addition to providing valuable insights into public perceptions of parking services, the findings can also be utilized by relevant stakeholders to improve service quality and formulate more appropriate policies. Furthermore, this study contributes to the development of sentiment analysis methods, particularly in the context of the Indonesian language and specific social issues such as parking services, by
providing a more in-depth error analysis of the model’s classification outcomes.