ANALISIS SENTIMEN KINERJA TENAGA MEDIS INDONESIA MENGGUNAKAN MODELING ROBERTA DAN METODE MACHINE LEARNING

Abstract

The development of information technology has led to the emergence of various social media applications, such as X, which allow users to share information and opinions. However, social media can also function as a platform for the spread of hoaxes and hate speech. One of the challenges faced is determining whether user comments are negative, neutral, or positive through sentiment analysis. This study aims to compare the performance of various classification algorithms in sentiment analysis of medical personnel services in X, using the Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbors (KNN), and Naïve Bayes (NB) algorithms with RoBERTa model-based labeling. Data was collected through the tweet-harvest library and processed using the Python programming language. The results showed a significant increase in accuracy, with the model able to classify public opinion into positive, negative, or neutral categories. The SVM model achieved the highest accuracy of 91.8%, outperforming other models such as Random Forest and Naïve Bayes, and provided insight into sentiment towards government health services. These findings provide valuable insights for policymakers in improving the provision of health services and managing public perceptions of medical personnel.