Analisis Sentimen Dosen Terhadap Kebijakan Tunjangan Kinerja (Tukin) Di Perguruan Tinggi Negeri Menggunakan Algoritma Naive Bayes

Abstract

This study aims to analyze lecturers' sentiments toward the performance allowance (Tukin) policy at public universities by applying the Naive Bayes algorithm. The data used consisted of 607 records collected through data crawling using Tweet-Harvest; however, after preprocessing, only 259 data points remained. For the initial labeling process, a simple sentiment analysis was performed to classify the data as either positive or negative sentiment. The labeling results showed a dominance of negative sentiment at 94.57%, while positive sentiment accounted for only 5.43%, reflecting a high level of dissatisfaction among lecturers regarding the policy's implementation. The Naive Bayes model was trained using a data split of 80% for training and 20% for testing. The model evaluation demonstrated an accuracy of 83.78%, with an F1-Score of 90.32% for negative sentiment and 50.00% for positive sentiment. The confusion matrix indicated that the prediction of negative sentiments was highly accurate, while the prediction of positive sentiments still needs improvement due to the imbalance in the data distribution. The results of this study indicate that the Naive Bayes algorithm is quite effective in classifying lecturers' opinions regarding the Tukin policy, particularly in detecting negative sentiments.